News for hackers. Stuff that matters.
Re:Top model scores may be skewed by Git history leaks in SW... by lieret on Thursday September 11, 2025 @08:12PM (#45217428)
[On swe-bench team] We read and analyzed a lot of trajectories but seems like only recently models have started to exploit this in a small fraction of instances. But yes, clearly shouldn't have happened (and is now fixed in the new container versions).

There are no comments on this comment yet.