Revision History
MATH-500 · 12 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.
- Update: regenerate results chart from the fact-checked results table
Version 12 · Nov 5, 2026, 07:11 AM
- Correction: Corrected MATH-500 results rows that actually used the full MATH test set (GPT-4 Turbo, GPT-4o) or the wrong model (o1-preview vs o1) or had no source (Gemini 2.5 Pro), the o1 94.8%/96.4% explanation, the MATH paper's 6.9% attribution, two misquoted Vals AI passages and a misdated Gemini 3 Pro score, Artificial Analysis's current treatment of MATH-500, and stale FrontierMath figures (plus 18 minor fixes)
Version 11 · Nov 5, 2026, 07:11 AM
- Correction: Sarvam-105B was released in 2026, not 2025
Version 8 · Oct 27, 2026, 10:35 AM
- Improve internal links (2026-10-13)
Version 7 · Oct 13, 2026, 10:04 AM
- Added 6 contextual internal links
Version 6 · Jul 23, 2026, 03:31 PM