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

Massive Multitask Language Understanding on MMLU-Pro (Accuracy)

80.6Accuracy (MMLU-Pro)

MIPROv2

40.14450.64761.1571.653Nov 25, 2025Dec 13, 2025Jan 1, 2026Jan 20, 2026Feb 7, 2026Feb 26, 2026Mar 17, 2026
Updated 16d ago

Evaluation Results

MethodLinks
2025.11
80.6
2025.11
80.1
2025.11
79.7
2026.03
78.6
2025.11
78.4
2025.11
77.7
2026.03
77.3
2025.11
77.1
2025.11
76.5
2025.11
76.3
2025.11
76.2
2025.11
76.1
2025.11
75.4
2025.11
75.3
2025.11
75.3
2025.11
71.1
2025.11
70.3
2025.11
68.7
2025.11
68.6
2025.11
68.6
2025.11
68.5
2025.11
68.5
2025.11
67.6
2025.11
66.2
2025.11
66.1
2025.11
64.7
2026.03
63.1
2025.11
62.7
2025.11
62.2
2025.11
60.7
2026.03
57.3
2025.11
56.5
2026.03
56
2026.03
49.2
2026.03
48.2
2026.03
47.1
2025.11
44.9
2025.11
41.7