👋 Hello there, I’m Runyu (Cathy)!

Welcome to my website! I am Runyu (Cathy) Zhang, a sixth year Ph.D. candidate in School of Engineering and Applied Science at Harvard University working with Prof. Na Li. Prior to pursuing Ph.D., I earned a B.S. degree in Scientific and Engineering Computing in the Mathematics department at Peking University in 2019.

My research interest lies in learning, control and decision making in multi-agent systems. The high-level objective is to design scalable, efficient and provable learning/control algorithms for multi-agent systems under challenges such as communication constraints, strategic behavior and model uncertainty. My research sits at the intersection of various domains, utilizing tools from reinforcement learning (RL), game theory, control theory and optimization. As an ultimate goal, my research is dedicated to providing both theoretical insights and engineering tools for AI-enabled multi-agent societal systems design and operation. For more detailed research projects and research interests please refer to the research tab and publication tab.

Recent Updates

Aug 2024 I'm very honored to be selected as rising star in EECS 2024! June 2024 Check out our new paper: Equilibrium Selection for Multi-agent Reinforcement Learning! Apr 2024 Our paper Multi-Agent Coverage Control with Transient Behavior Consideration is accepted to L4DC. The paper provides a provably efficient algorithm for coverage control of multi-robot systems in unknown environments. Feel free to check it out! Feb 2024 Our paper Gradient play in stochastic games: stationary points, convergence, and sample complexity is accepted to Transaction of Automatic Control (TAC)! Jan 2024 Our paper Soft Robust MDPs and Risk-Sensitive MDPs: Equivalence, Policy Gradient, and Sample Complexity is accepted to ICLR. Jan 2024 Two papers are accepted to the American Control Conference (ACC)! Check them out here and here. July 2023 Our paper Markov Games with Decoupled Dynamics: Price of Anarchy and Sample Complexity is accepted to the Conference on Decision and Control (CDC)! Jan 2023 Our paper (one of my favorite works!) On the Optimal Control of Network LQR with Spatially-exponential Decaying Structure is accepted to the American Control Conference (ACC)! We are in preparation for a journal version as well. Feel free to check it out!! Sep 2022 Two papers on multi-agent RL is accepted to NeurIPS! See you in New Orleans! Check them out here (Markov potential games) and here (two-player zero-sum games).