Are AI Evals Broken? Anthropic/NYU’s Pavel Izmailov on LLM Evaluation, Reasoning & “Alien” Behavior

Explore this episode

Prepare a transcript and summary so you can explore the episode and ask questions about what was said.