Baoxiang Wang bio photo

Baoxiang Wang

Associate Professor, The Chinese University of Hong Kong, Shenzhen

Email Google Scholar LinkedIn Github

I am an associate professor at the School of Data Science, The Chinese University of Hong Kong, Shenzhen. I was an assistant professor from Aug 2020 to Jun 2026. I work on reinforcement learning and game theory. My recent works focus on LLM-RL, learning in games, and information design among multiple agents.

I obtained my Ph.D. in Computer Science and Engineering from The Chinese University of Hong Kong in 2020, fortunate to be advised by Siu On Chan and Andrej Bogdanov. Before that, I obtained my B.E. in Information Security from Shanghai Jiao Tong University in 2014, under Hai Zhao.

I am currently a visiting researcher at Vector Institute in Toronto, since Dec 2025, with Yaoliang Yu. I held the same position there during Jan-Aug 2024, and was at University of Waterloo as a visiting assistant professor during May-Sep 2023, with Pascal Poupart.

Education

2014 - 2020 Ph.D. The Chinese University of Hong Kong

2010 - 2014 B.E. Shanghai Jiao Tong University

Research directions

  • LLM-RL. Policy optimization for long-horizon LLM-RL, including variance reduction and trust-region methods, together with scalable exploration and the convergence analysis of the underlying algorithms.
  • Learning in games. Uncoupled and no-regret learning dynamics that converge to equilibria in monotone, potential, extensive-form, and mean-field games, mostly under bandit feedback.
  • Information design and mechanism design. Bayesian persuasion, communication, and incentive mechanisms among multiple agents, and learning-based and LLM-based methods for solving them.
  • LLM agents in games. Negotiation, cooperation, and debate among self-interested agents, LLM agents in particular, and understanding equilibria of games such as poker with natural language.
  • Ad-hoc topics in learning theory and game theory that I find interesting.

Prospective students

I am recruiting PhD students for Fall 2027 enrollment. The degree will be issued by CUHK for PhD in Computer Science or PhD in Data Science, with a mark that the study is completed on the Shenzhen campus. The programs are similar to the ones in Hong Kong. See the program information.

For candidates, I care mostly about their math background and their motivation for working on new and exploratory problems in reinforcement learning and game theory. Students are required to use AI tools to advance their research.

Contact

  • Office: 503 DY, CUHK Shenzhen. Shenzhen, 518172
  • Phone: +86 755 235-17020