About Me

I am a Professor at the Artificial Intelligence Innovation and Incubation Institute (AI³) at Fudan University and an AI Scientist at the Shanghai Academy of AI for Science (SAIS).

Prior to my current roles, I gained extensive industry experience serving as the Head of Post-Training at Infinity, as well as holding researcher positions at Ant Group and ByteDance. In 2018, I was a postdoctoral fellow at the Technion – Israel Institute of Technology, hosted by Prof. Shie Mannor. I received my Ph.D. degree from the National University of Singapore (NUS) in 2017, advised by Prof. Huan Xu. During my doctoral studies, I also spent a year (2016-2017) as a visiting scholar at the H. Milton Stewart School of Industrial and Systems Engineering (ISyE) at Georgia Tech.

CHAOS Lab

CHAOS Lab (Collaborative & Hierarchical Agents for Open-ended Science) develops general-purpose autonomous agents for reasoning, strategic interaction, and scientific discovery.

Research Interests

My long-term goal is to develop “general-purpose autonomous agents” capable of deep reasoning, strategic interaction, and independent scientific discovery. My current research bridges reinforcement learning, foundation models, and AI for Science, focusing on the following key directions:

• RL for Foundation Models: Advancing the post-training paradigms for LLMs, VLMs, and VLA models, with a specific focus on solving core challenges in deep exploration and critic learning.
• LLM-Empowered Multi-Agent RL: Designing scalable MARL frameworks that leverage the reasoning capabilities of Large Language Models to achieve complex multi-agent coordination and strategic decision-making.
• Learn to Discover (AI4S): Building autonomous scientific discovery agents that can actively interact with their environments to automate research workflows and close the “dry-wet” experimental loop.

Group Members

Ph.D. Students

• Zhijian Zhou — Reinforcement learning, multi-agent RL for LLMs, and LLM RL infrastructure.
• Xuan Zhang — Deep exploration in RL for LLMs and world models.
• Yuchen Liu — Memory mechanisms in large language models.

Master Students

• Kangcheng Xiao — Automated research agents.
• Ze He — Text-to-video generation.

Undergraduate Students

• Tongjin Zou — Automated research agents.

Alumni

• Long Li — RL for agents.

Open Positions

• Fudan University (Ph.D. / Master / Undergrad): I am seeking highly motivated full-time and part-time students to join my group at Fudan. If you have a strong background in computer science or mathematics and are passionate about RL, LLMs, or multi-agent systems, please reach out.
• SAIS Internships: I am recruiting on-site research interns to work with me at the Shanghai Academy of AI for Science (SAIS). This is an excellent opportunity to dive into cutting-edge problems in foundation models and AI4S.
• Remote Scholars & Collaborations: I am welcoming remote interns, visiting scholars, and broad academic or industrial collaborations. I am open to discussing self-funded or externally funded opportunities without strong restrictions on topics.