I am a Ph.D. student at Institute of Automation, Chinese Academy of Sciences (CASIA) and Zhongguancun Academy (ZGCA). Before that, I received my B.E. degree in Computer Science and Technology with overall ranking 1/449 (0.22%) at the School of Computer Science from Beijing University of Posts and Telecommunications (BUPT) in Jun. 2024. I am grateful to be growing up and studying with my twin brother Xuzhao Li (M.S. Student at BIT), which is a truly unique and special experience for me.
- Multi-modal Adaptive Agentic RL and OPD: Empower agents through reinforcement learning and on-policy distillation to autonomously interact with complex, dynamic environments and carry out multi-turn, long-horizon decision-making tasks, enabling them to better process text-image and video information.
- Streaming Video Understanding Memory Modeling: Optimize the dense memory compression, memory retrieval triggering mechanism and memory window management capabilities of streaming video understanding models.
- Fine-grained Evaluation of Multi-modal Reasoning: Build benchmarks targeting the model’s reasoning process to conduct fine-grained evaluation and diagnosis of logical consistency, reasoning effectiveness and the distribution of key information.