-
코딩 평가에서 신호와 노이즈 구분하기
Separating signal from noise in coding evaluations
A new analysis from OpenAI reveals issues in SWE-Bench Pro, a popular coding benchmark, raising concerns about reliability and accuracy in evaluating AI models.
-
신뢰할 수 있는 제3자 평가를 위한 공유 플레이북
A shared playbook for trustworthy third party evaluations | OpenAI
OpenAI shares guidance on third-party AI evaluations, covering how to assess model capabilities, safeguards, and validity for frontier systems.