PhD student, Stanford University
1 paper at NeurIPS 2025
A memory framework for video world models that improves long-term consistency