I am a second-year Ph.D. student in Computer Science Department, UCLA, advised by Prof. Song-Chun Zhu. As a member of Center for Vision, Cognition, Learning and Autonomy (VCLA), I am also actively advised by Prof. Ying Nian Wu. I obtained BEng degree in Computer Science from The Hong Kong University of Science and Technology (HKUST), advised by Prof. Fangzhen Lin.
My current research focuses on learning generalizable representations of i) core cognitive knowledge such as objects and physics, ii) high-level concepts such as categories and abstract relations, and iii) task-oriented abstraction such as options, affordance and utility, for out-of-distribution reasoning and generalized planning. The ultimate goal is to endow embodied agents with a computational framework to develop environment-invariant general knowledge from interaction and communication. I draw inspiration broadly from classical AI, cognitive science, and computational neuroscience, etc. Previously, I studied (Inverse) Reinforcement Learning and Multi-agent Systems. My works on the SNAS series pioneer the field of Differentiable Neural Architecture Search.