Benchmark

MMLU-Redux

Error-corrected MMLU subset; reported mainly by open-weight cards.

Released
2024-06-06
Categories
knowledge
Openness
unknown
Source
model_reports
Reported scores
4

Reported scores

model_reports

ModelOrganizationReported valueReportedEvidence
DeepSeek-R1DeepSeek92.92025-01-22evidence
DeepSeek-V3DeepSeek89.12024-12-27evidence
Qwen3-235B-A22B (Thinking)Qwen92.72025-05-14evidence
Qwen3.5-397B-A17BQwen94.92026-02-16evidence

Scores are partitioned by the source that reported them and are never merged into a single cross-source ranking, because the sources measure different things and say so.