MMLU
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
MMLU-Pro
84.3Accuracy
4
MMLU-PRO
84.6Score
4
MMLU
89.28Accuracy
4
MMLU redux
0.269Accuracy
4
MMLU STEM (out-of-distribution)
93.3Accuracy
4
MMLU
64.47MMLU Pass@1 Accuracy
4
MMLU-Pro
75.3MMLU-Pro Accuracy
4
MMLU College Physics 1.0 (test)
95.1Success Rate
4
MMLU Machine Learning 1.0 (test)
88.4Accuracy
4
MMLU College Medicine
89Accuracy
4
MMLU College Biology
95Accuracy
4
MMLU Anatomy
89Accuracy
4
MMLU Medical Genetics
96Accuracy
4
MMLU PM
91.2Accuracy
4
MMLU (dev)
86.4Accuracy
4
MMLU-Pro Economics discipline
10.01Mean EMR
4
MMLU
1.95MAE
3
MMLU (test)
7.48Anatomy Accuracy
3
MMLU-Pro
56.7Health Score
3
MMLU
0.1Drift (pp)
3
MMLU-Redux
90.8Exact Match (EM)
3
MMLU professional law
2.71NFE Speedup
3
MMLU moral scenarios
2.82NFE Speedup
3
MMLU clinical knowledge
3.18NFE Speedup
3
MMLU
0.655Kendall's Tau Correlation
3