General Knowledge
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
44.4MMLU-Pro Score (0-shot)
9
Apr 10, 2026
55.88MMLU
9
Feb 26, 2026
78.73CoT EM
8
Jun 16, 2026
44.06MMLU EL (General Knowledge) Accuracy
8
Mar 6, 2026
11,691.83Avg Input Context Length (tokens)
8
Feb 26, 2026
78.12MMLU* Accuracy (high-perf.)
7
Jun 16, 2026
82.7MMLU Accuracy
7
Feb 26, 2026
57.3KMMLU
7
Feb 26, 2026
37.4MMLU General Knowledge Drop in Utility (%)
6
Apr 14, 2026
82.39MMLU Stem GK Accuracy
6
Mar 17, 2026
70.7Solution Rate
6
Feb 26, 2026
92.9Exact Match
6
Feb 26, 2026
91.8EM
6
Feb 26, 2026
88.7Accuracy@8
5
Jul 3, 2026
50Exact Match (EM)
5
Jun 16, 2026
68.18Pass@3
4
Mar 24, 2026
61.8pass@1
4
Feb 26, 2026
18.6Accuracy
4
Feb 26, 2026
60CoT EM
3
Apr 15, 2026
48Winrate
1
Jun 23, 2026
36Winrate
1
Jun 23, 2026
1.3MMLU General Knowledge Delta (pp)
1
Jun 15, 2026