I am a final-year Ph.D. candidate and Interdisciplinary Graduate Fellow at Stanford University, advised by Mac Schwager. I am currently a Student Researcher at Google DeepMind. I am also affiliated with the Stanford Center for AI Safety. Before Stanford, I received my M.Phil. from the Multimedia Laboratory at The Chinese University of Hong Kong, advised by Bolei Zhou, and my B.S. in Computer Science from Shanghai Jiao Tong University, advised by Cewu Lu. Previously, I spent time at Meta Superintelligence Lab and Microsoft Research.
I am a final-year Ph.D. candidate and Interdisciplinary Graduate Fellow at Stanford University, working with Mac Schwager. I am a student researcher at Google DeepMind. I am also affiliated with the Stanford Center for AI Safety. I have spent time at Meta Superintelligence Lab and Microsoft Research.
My research advances general-purpose physical intelligence by developing AI systems that build predictive multimodal world representations, reason efficiently over long horizons, and act reliably under physical constraints and uncertainty.
To this end, I develop principled foundation-model architectures, predictive multimodal world models, and compositional learning methods for long-horizon agentic systems. My goal is to create efficient, adaptable agents that learn from broad experience while remaining grounded in physical dynamics, uncertainty, and human needs.
Derive efficient mechanisms for attention, memory, routing, and normalization from their underlying mathematical structure.
Learn predictive representations of geometry, dynamics, semantics, and uncertainty for reasoning and control.
Compose reusable skills across temporal scales while controlling error accumulation and recovering from failure.
