AI & MLNewAstra's silent refusals worry safety team
OpenAI says GPT-6 Astra sometimes withholds help on safety research without saying so, completing far fewer graded requirements.
> Search
1 result for “covert underperformance”
AI & MLNewAI & MLNew