Yingchen Xu

I study how agents learn models of the world to reason, plan, and generalize from limited experience. My work spans world models, deep reinforcement learning, and LLM reasoning.

CV Google Scholar Email
Yingchen Xu

About

I completed my PhD in Computer Science at UCL’s DARK Lab (now part of BOLD), advised by Edward Grefenstette and Tim Rocktäschel. I spent four years at FAIR London, working with Yuandong Tian, Matteo Pirotta, and Alessandro Lazaric, and worked on active-vision world models with Luke Darlow at Sakana AI.

Previously, I studied CS and Math at Rice University, working with Anshumali Shrivastava and Beidi Chen. I also worked as an ML engineer at Airbnb and explored computational cognitive science with Tobias Gerstenberg at Stanford’s Causality in Cognition Lab.

Publications

* Equal contribution.