Zhuokai Zhao· zhuokaiz·· 13 天前AI 评分26
Reflection AI 模型开源基准NR争议
this looks great! can't wait for @reflection_ai to release the checkpoint! but the second thing that came to my mind was, whoever signed ...
AI 导读
这看起来很棒！迫不及待想等 @reflection_ai 发布 checkpoint！但我脑海中闪过的第二件事是，批准这张图表的人，你们怎么能接受在 SWE-bench Verified 这类基准测试中，将开源模型标记为 NR？
来源：Zhuokai Zhao · x.com