Revision History
SWE-bench Verified · 8 revisions
Sizes are character counts of the article source. The signed number is the change from the previous revision.
Recent edit summaries
Detailed summaries recorded by editors. Generic maintenance summaries are omitted here; every recorded revision remains below. Dates describe the edit, not necessarily the event it covers.
- Correction: Corrected a fabricated OpenAI quote and invented annotator-pool details, the misdescribed annotation rubric, wrong repository shares, unsupported leaderboard rows (Devin, o1, GPT-5.3 Codex, Claude Sonnet 4), an invented third 'saturation' finding attributed to OpenAI's deprecation post, misquoted test-weakness study figures and SWE-Lancer comparisons, and updated the stale top score to Claude Opus 5's self-reported 96.0% (plus 23 minor fixes)
Version 8 · Nov 1, 2026, 11:40 AM
- Added 6 contextual internal links
Version 7 · Jul 23, 2026, 03:03 PM