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

Multi-task Language Understanding on MMLU (Top-1 agreement)

74.7Top-1 Accuracy

HARP-C

-2.67617.41237.557.588Apr 17, 2025Jun 30, 2025Sep 13, 2025Nov 26, 2025Feb 9, 2026Apr 24, 2026Jul 8, 2026
Updated 16d ago

Evaluation Results

MethodLinks
2026.06
74.7
2026.06
74.2
2026.06
66.5
2026.07
65.25
2026.07
63.22
2026.06
63.2
2026.06
62.7
2026.06
62.7
2026.06
62.4
2026.06
62.2
2026.06
61.8
2026.06
61.4
2026.06
61.3
2026.06
61.3
2026.07
61.3
2026.07
61.2
2026.06
60.5
2026.07
60
2026.06
59.4
2026.07
59.17
2026.06
58.6
2026.07
57.6
2026.06
57.3
2026.06
56.5
2026.07
54.88
2026.06
54.1
2026.06
54
2026.06
47.5
2026.07
46.9
2026.06
45.5
2025.04
36.47
2025.04
35.4
2025.04
33.28
2025.04
33.03
2025.04
32.98
2025.04
31.92
2025.04
31.6
2026.07
22.94
2026.07
22.93
2026.07
22.91
2026.07
22.88
2026.06
13.8
2026.06
13.6
2026.06
8.3
2026.06
2.2
2026.06
0.3