SWE-Bench Pro Faces Backlash Over Limited Repository Diversity

The AI benchmark SWE-Bench Pro is facing criticism for its lack of representativeness, as users point out it only contains 11 repositories and 731 tasks. Commenters argue this limited diversity makes it an unreliable metric for evaluating true coding capabilities.

2026-07-22 ~ 2026-07-22 · 2 related posts