berkeley-icon-talk

Excited to give a talk at UC Berkeley ICON Lab on how to build scalable inference-time policy steering with world modeling and verification!