Xuehai Pan (/ʃwɛˈhaɪ pæn/, 潘学海 in Mandarin, Xuehai.Pan@outlook.com) holds a Ph.D. degree in Applied Computer Science from Peking University. His research interests lie in the intersection of Reinforcement Learning, Multi-Agent Systems, Distributed Computing, and Autonomous Agents, with a focus on developing scalable and automated algorithms and exploring their theoretical and practical aspects. He has a solid background in both research and engineering, having obtained a B.S. degree in Physics with honors and a B.S. degree in Computer Science (double major) from Peking University before pursuing his Ph.D. degree. His academic journey is embellished with achievements such as winning gold medals in the Chinese Physics Olympiad (CPhO) and the Asian Physics Olympiad (APhO) during high school.
Xuehai now works on building autonomous agents — systems that can reason, plan, use tools, and collaborate to carry out complex real-world tasks with minimal human oversight — while keeping them trustworthy and aligned with human intentions and values. He is equally interested in using agents to build the next generation of agents — automating data synthesis, red teaming, evaluation, and evolutionary training through multi-agent interaction and self-play. The ultimate goal is to automate everything — not only the real-world tasks agents carry out, but the very process of creating, training, aligning, and governing the agents themselves, so the entire loop runs autonomously and at scale.
Beyond academia, Xuehai is an open-source enthusiast and an active contributor to influential projects such as PyTorch, CPython, Ray, Transformers, DeepSpeed, Gymnasium (formerly OpenAI Gym), PyBind11 (C++ bindings for Python), PyO3 (Rust bindings for Python), Conda, Homebrew, etc. He enjoys dedicating his spare time to helping people and sharing knowledge in the community, further enriching his impact beyond his research pursuits.
