Yuxuan Zhang

Hi, I’m an M.S. student in Computer Science at UC San Diego (UCSD), fortunately advised by Prof. Hao Zhang and Prof. Haojian Jin.

Before that, I earned my B.S. in Data Science from UCSD’s Halıcıoğlu Data Science Institute (HDSI), graduating summa cum laude (top 2 %).

I am interested in AI Agents: 50% on building AI agents and 50% on agentic RL training.

Highlights

Large‑Model Benchmark: LMGame‑Bench  (2025)

Agentic Multi-Turn RL Training: LMGame‑Train (2025)

AI Gameplay Demo: Ace Attorney Gameplay (2025)

Publications

1. LMGame‑Bench: How Good are LLMs at Playing Games? arXiv 2505.15146
   NeurIPS ’25 — submitted
   Lanxiang Hu, Mingjia Huo, Yuxuan Zhang, Haoyang Yu, Eric P. Xing, Ion Stoica, Tajana Rosing, Haojian Jin, Hao Zhang

2. General Modular Harness for LLM Agents in Multi‑Turn Gaming Environments arXiv 2507.11633
   ICML ’25 Workshop
   Yuxuan Zhang, Haoyang Yu, Lanxiang Hu, Haojian Jin, Hao Zhang