Updated daily at 05:00 UTC
Benchmark
MMLU-ProX
Measures: MMLU-Pro's 10-option reasoning questions translated into 29 typologically diverse languages (11,829 parallel questions per language; 658-question Lite set); accuracy averaged over languages.
- Maintainer
- University of Tokyo and collaborators (Xuan et al.)
- Introduced
- 2025
- Status
- Active
- Saturation
- Near-ceiling
- Typical provenance
- Self-reported
- In scoring set
- No
Leaderboard
Independent and third-party results only.
No independent or third-party results yet. Show all results