About me
I am a first-year Ph.D. student at MMLab, The Chinese University of Hong Kong (CUHK), after receiving my B.E. degree in 2026. 我于 2026 年获得工学学士学位，现为香港中文大学（CUHK） MMLab 的一年级博士生。
Research
- Agents: Reasoning and planning, reinforcement learning, memory and tool use, data and training, inference systems, and evaluation. Toward self-improving agents.智能体：推理与规划、强化学习、记忆与工具使用、数据与训练、推理系统和评估。面向持续自我改进的智能体。
- Physical AI & World Models: Multimodal representations, generative and action-conditioned world models, simulation, embodied learning, planning, and control. Toward general-purpose physical intelligence.物理智能与世界模型：多模态表征、生成式与动作条件世界模型、模拟、具身学习、规划与控制。面向通用物理智能。
Updates
-
Reward Is Not a Universal Interface for Generative Reinforcement Learning was accepted to NeurIPS 2026.
Reward Is Not a Universal Interface for Generative Reinforcement Learning 被 NeurIPS 2026 接收。