OpenAI News·· 2026-02-23AI 评分37
OpenAI 停止评估 SWE-bench Verified 的原因
Why we no longer evaluate SWE-bench Verified
AI 导读
OpenAI 分析发现,SWE-bench Verified 评测基准存在测试缺陷和训练泄露问题,因此不再进行评估,并推荐使用 SWE-bench Pro。
原文
SWE-bench Verified is increasingly contaminated and mismeasures frontier coding progress. Our analysis shows flawed tests and training leakage. We recommend SWE-bench Pro.
来源:OpenAI News · openai.com