I am an Embodied Intelligence Algorithm Researcher at AgiBot’s Embodied Intelligence Research Center (具身研究中心). I received my Ph.D. in Computational Mathematics from Peking University in 2026, advised by Prof. Bin Dong.

My research sits at the intersection of mathematics and applied artificial intelligence. I am broadly interested in embodied intelligence and robot learning, large language models and synthetic data, reinforcement learning, and inverse problems.

Research Interests

• Embodied intelligence and robot learning
• Large language models and synthetic data
• Reinforcement learning
• Inverse problems and mathematical modeling

Selected Publications

• Learning while Deploying: Fleet-Scale Reinforcement Learning for Generalist Robot Policies
  arXiv, 2026. Paper
• Spend Wisely: Maximizing Post-Training Gains in Iterative Synthetic Data Bootstrapping
  NeurIPS 2025 Spotlight. Paper
• Beyond Model Collapse: Scaling Up with Synthesized Data Requires Reinforcement
  ICLR 2025. Paper
• L2SR: Learning to Sample and Reconstruct for Accelerated MRI via Reinforcement Learning
  Inverse Problems, 2024. Paper