My name is Viktor Volkov, and I am an aspiring researcher in the field of ML/DL.
My current research revolves around LLM-based AI agents and Multi-Agent Systems.
I also work with Neural Architecture Search (NAS) and Evolutionary Algorithms.

• 🤖 LLM Agents / Multi-Agent Systems
• 🧬 Neural Architecture Search
• 🌱 Evolutionary Algorithms

• ML/DL: PyTorch, TorchVision, scikit-learn, Optuna, HuggingFace Transformers, LightGBM
• LLM & Multi-Agent: Transformers, LangChain, LangGraph, Ollama Python
• Data Processing: Pandas, NumPy, Matplotlib, Seaborn
• DevOps: Git, Docker, Docker-compose
• Databases: PostgreSQL, SQLAlchemy, ChromaDB, Redis
• Backend: FastAPI

• ⚔️ HeroBench: A Benchmark for Long-Horizon Planning and Structured Reasoning in Virtual Worlds - A1 Agent developer.
• 🧬 BlockNAS – Simple NAS framework for joint NAS and HPO optimization.
• 📄 Simplified reproduction of the article "Can Large Language Model Agents Simulate Human Trust Behavior?" – measuring trust in small local LLMs in the Dictator Game. Project completed for the summer school "Summer with AIRI 2025".

💬 You can also reach out via the other methods on my website.