I am a second year M.S. in Data Science (thesis track) at UC San Diego.
I’m working as a research assistant in Professor Biwei Huang’s lab, where I’m collaborating with and advised by Fan Feng (postdoc@UCSD/CMU) and Caleb Chuck (postdoc@Princeton). Previously, I received my bachelor’s degree in Data Science from East China Normal University, Shanghai.
Research interests
-
Unsupervised Reinforcement Learning
- How can agents discover diverse, controllable skills through exploration without task-specific rewards?
- How can we transfer skills above to downstream tasks?
-
Robot Learning
- How can we learn behavior models/representations that prioritize control over visual reconstruction?
- How can robots assess their own capabilities and continually evolve?
- How can robots transfer skills to control unseen objects with little or no task-specific data?
Publications
Density Matters: Normalized Successor Measures for Zero-Shot Control
Preprint, 2026
Project Page Paper (coming soon)
Selected Projects
-
Three Faces of the No-Free-Lunch Theorem
- First Lean4 Proof of Wolpert–Macready NFL, and Batch Stochastic NFL
-
Unsupervised Skill Discovery in Continuous Control
- Studied and Implemented intrinsic rewards in MuJoCo control environments
-
Deep Learning Systems
- AutoDiff, Transformer; Matmul+Tiled/KernelOpt; MoE, Collective Communication