Hi! I’m a 3rd year undergraduate Researcher and Engineer at UC Berkeley studying Electrical Engineering and Computer Science (EECS).

I’m broadly interested in Reinforcement Learning and Robot Learning. Currently, I work on Policy Extraction and offline-to-online RL at BAIR, where I’m fortunate to be advised by professor Sergey Levine. I’ve also had the chance to work on Test-Time methods for RL as an OpenAI research fellow, and RL for VLAs with the GEAR team at NVIDIA. Previously, I’ve also worked on engineering LLM Infra with the ChipNemo team at NVIDIA, and Energy Validation Infra for Robotaxi Charging at Tesla. Outside research, I build machine learning applications with my friends at Cal Launchpad.

If you’d like to chat, please reach out at [andypeng at berkeley dot edu].

---

Publications

Self-improving Vision-Language-Action models with data generation via Residual RL Wenli Xiao*, Haotian Lin*, Andy Peng, Haoru Xue, Tairan He, Yuqi Xie, Fengyuan Hu, Jimmy Wu, Zhengyi Luo, Linxi "Jim" Fan, Guanya Shi, Yuke Zhu, International Conference on Learning Representations (ICLR), 2026 [paper] [website]

Efficient Online Reinforcement Learning Fine-Tuning Need Not Retain Offline Data Zhiyuan Zhou*, Andy Peng*, Qiyang Li, Sergey Levine, Aviral Kumar International Conference on Learning Representations (ICLR), 2025 [paper] [website] [code]