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

Multilingual Language Understanding on MMMLU (test)

66.88Accuracy

MSD (Off-Policy)

15.680828.972942.26555.5571May 3, 2026
Updated 2mo ago

Evaluation Results

MethodLinks
2026.05
66.88
2026.05
66.76
2026.05
65.49
2026.05
65.45
2026.05
65.26
2026.05
65.23
2026.05
65.23
2026.05
65.15
2026.05
65.07
2026.05
64.78
2026.05
64.65
2026.05
64.61
2026.05
59.22
2026.05
58.8
2026.05
58.26
2026.05
58.25
2026.05
58.17
2026.05
58.06
2026.05
58.04
2026.05
57.92
2026.05
57.9
2026.05
57.85
2026.05
57.47
2026.05
57.34
2026.05
57.16
2026.05
50.55
2026.05
48.17
2026.05
48.16
2026.05
47.87
2026.05
46.57
2026.05
44.01
2026.05
42.26
2026.05
41.53
2026.05
35.31
2026.05
34.88
2026.05
34.28
2026.05
33.79
2026.05
32.05
2026.05
32.01
2026.05
31.97
2026.05
31.96
2026.05
31.91
2026.05
31.86
2026.05
31.84
2026.05
31.61
2026.05
31.49
2026.05
31.33
2026.05
31.2
2026.05
27.11
2026.05
24
2026.05
22.86
2026.05
17.65