Apollo Research 与 OpenAI 启动嵌入式评估试点,调查模型是否试图破坏对自身的监督
AI 导读
Apollo Research 与 OpenAI 启动嵌入式评估试点,六名评估员将入驻 OpenAI 工作四个月,调查模型是否会试图破坏对自身的监督与控制。
正文 · 原文
We're launching a pilot towards embedded evaluations with @OpenAI.
Six evaluators from Apollo Research will work within OpenAI for four months, investigating whether models attempt to undermine their own oversight and control.