Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

General Language Capability on MMLU, GSM8K, HEval, IFEval, MTBench

76.4MMLU Accuracy

Baseline

68.18470.31772.4574.583Jul 1, 2026
Updated 23d ago

Evaluation Results

MethodLinks
2026.07
76.480.86453.682.471.4
2026.07
76.482.46449.879.170.3
2026.07
76.483.26451.880.271.1
2026.07
76.379.664.60.41046.2
2026.07
76.380.468.35182.671.7
2026.07
76.280.463.451.279.170.1
2026.07
768059.133.166.663
2026.07
74.576.46451.882.569.8
2026.07
7075.659.146.870.964.5
2026.07
69.974.462.252.175.566.8
2026.07
69.976.860.445.770.664.7
2026.07
69.973.659.85076.666
2026.07
69.874.861.649.277.466.6
2026.07
69.875.66151.277.267
2026.07
69.76661.626.851.655.1
2026.07
68.5585041.96155.9