A shared playbook for trustworthy third party evaluations
中文摘要
OpenAI发布第三方AI评估指南，涵盖前沿系统能力、安全防护及有效性的评估方法。
English Summary
OpenAI released guidance on third-party AI evaluations, covering how to assess capabilities, safeguards, and validity for frontier systems.
Original Excerpt
OpenAI shares guidance on third-party AI evaluations, covering how to assess model capabilities, safeguards, and validity for frontier systems.