PhD Student in CS, UMass
I am a first-year PhD student in CS at UMass Amherst, advised by Prof. Chuang Gan. My research focuses on robot learning, with the goal of building general-purpose robots that can learn diverse skills and adapt to unfamiliar tasks and environments. Prior to my PhD, I was a full-time research scientist at Genesis AI. I received my master’s degree from UMass Amherst and my bachelor’s degree from Fudan University.
* Equal contribution
arXiv, 2026
Simulation-initialized generated videos become executable dexterous robot demonstrations through motion reconstruction, retargeting, and residual reinforcement learning.
CoRL 2026
Image editing and simulator priors turn semantic goals into physically executable supervision for robot manipulation.
ICLR 2026
LLM agents work with a physics solver to build complex 3D arrangements with realistic contact, support, and containment.
NeurIPS 2025
Vision-language models and physics simulation jointly design tools and optimize how robots use them.
CoRL 2025
Vision-language models convert rigid 3D meshes into articulated assets with functional parts and joints.
NeurIPS 2024
Hierarchical image inpainting creates detailed, interactive 3D environments for embodied AI.
ICML 2024
A generative world model connects 3D perception, language reasoning, and action through goal-image and point-cloud prediction.