John Schulman on Reinforcement Learning from Human Feedback
UC Berkeley EECS · Talk · 1 h 4 min · 2023 · YouTube title: John Schulman - Reinforcement Learning from Human Feedback: Progress and Challenges
Why it's here
Recommended in LLM threads as an authoritative explanation from a ChatGPT architect of why models hallucinate and why imitation is not enough.
What it covers
UC Berkeley EECS colloquium by OpenAI's John Schulman on reinforcement learning from human feedback. He covers why RL helps with truthfulness, hallucination, and the limits of behavior cloning.