Arash Ahmadian on Rethinking RLHF

af TalkRL: The Reinforcement Learning Podcast