Benchmark

CritPt

Measures: 70 unpublished, research-level physics challenges (11 subfields) written by 50+ active physicists, with guess-resistant answers (arrays, symbolic expressions, Python functions); scored as mean accuracy over 5 runs by the authors' grading server.

Maintainer
CritPt authors (Argonne National Laboratory and UIUC, 50+ contributing physicists); official leaderboard hosted by Artificial Analysis
Introduced
2025
Status
Active
Saturation
Headroom
Typical provenance
Independent
In scoring set
Yes

PaperOfficial leaderboard

Every result we hold, whoever ran it. Only independent and third-party rows count towards a score.

No results to show.

CritPt: leaderboard and what it measures · supermarket.dev