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

Utility Evaluation on MMLU and GSM8K

70.4MMLU Accuracy

PALETTE

48.45654.15359.8565.547May 22, 2026
Updated 8d ago

Evaluation Results

MethodLinks
2026.05
70.485.377.9
2026.05
70.186.578.3
2026.05
7086.278.1
2026.05
69.985.177.5
2026.05
69.985.677.8
2026.05
69.786.177.9
2026.05
68.381.775
2026.05
67.582.475
2026.05
66.880.473.6
2026.05
66.481.774.1
2026.05
64.879.772.3
2026.05
56.963.760.3
2026.05
56.963.760.3
2026.05
49.359.154.2
2026.05
49.359.154.2
2026.05
49.359.154.2