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

This episode hasn't been analyzed yet.

This episode is outside the automatic freshness window. Request analysis to use 1 included analysis request and generate the summary, transcript-aware context, and trade ideas.