MMLU
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
MMLU
82.06Accuracy
50
MMLU
71.91Average Accuracy
50
MMLU-Pro
90.1Accuracy
48
MMLU Phy
94.6ASR (Accuracy)
48
MMLU-Redux
95.2Accuracy
48
MMLU
100Accuracy
47
MMLU-Pro
59.64Mean Accuracy
47
MMLU
74.7Top-1 Accuracy
46
MMLU 5-shot
74.01Knowledge (5-shot) Score
46
MMLU
78.7Accuracy
46
MMLU
92Accuracy
46
MMLU
49.5MMLU Score
45
MMLU physics
90.1MDR
45
MMLU-Pro
81.4MMLU-Pro Score
44
MMLU STEM
86.6Accuracy
43
MMLU
82.1Accuracy
43
MMLU-STEM
87.4Accuracy
43
MMLU-PRO
51.91Accuracy (%)
42
MMLU Redux
84.13Accuracy
42
MMLU
82.57Accuracy
42
MMLU Pro.Med.
95.59Accuracy
42
MMLU
6.31CRR
42
MMLU
76.82MMLU Final Performance
42
MMLU o=1 Exact split
77.6Accuracy
42
MMLU Exact split, o=3
92.1Accuracy
42