2026
Embodied Playground
We test whether robotic agents can carry out a plan and adapt when an action changes the scene. Tasks range from rearranging a bookshelf to assembling an interlocking puzzle.
Chen Feng 陈枫
Hi there! You can also call me Winnie. I go by Winniechen online.
I'm a Ph.D. student at the University of Hong Kong, advised by Professor Yi Ma. My research goal is to build a universal robot to solve productivity challenges.
I graduated from the Institute for Interdisciplinary Information Sciences at Tsinghua University. I have conducted research at Tsinghua and was a visiting student working with robotics researchers at MIT.
Research interests
Agentic RoboticsRobot LearningDexterous ManipulationGenerative Simulation
Selected research
All publications →2026
2025
GenDexHand
In GenDexHand, we propose a generative pipeline for dexterous hand tasks that combines LLM/VLM refinement, motion planning, DoF constraints, and subtask decomposition to generate scalable, high-quality trajectory data.
2024
News
DexHoldem is on arXiv.
GenDexHand is on arXiv.
EvalTask is on arXiv.
Started my Ph.D. at the University of Hong Kong.
Say hello
If you're interested in chatting about robots, AI, research, cooking, Riot Games, the reason why we're here, or anything else, feel free to shoot me an email.
You can also find me in a game. 我们一起做一些有趣的事情吧！
Love yourselves guys, the future is ours.