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

Multi-task Language Understanding on MMLU (5-shot)

78.74Accuracy (5-shot)

Full Precision

20.791235.835650.8865.9244Feb 6, 2026
Updated 3mo ago

Evaluation Results

MethodLinks
2026.02
78.74
2026.02
76.88
2026.02
76.86
2026.02
74.51
2026.02
65.34
2026.02
59.35
2026.02
59.07
2026.02
58.82
2026.02
55.03
2026.02
54.56
2026.02
54.27
2026.02
46.45
2026.02
45.86
2026.02
45.55
2026.02
43.48
2026.02
43.38
2026.02
41.43
2026.02
39.79
2026.02
39.36
2026.02
37.66
2026.02
34.01
2026.02
29.37
2026.02
26.97
2026.02
26.63
2026.02
25.51
2026.02
25.47
2026.02
24.95
2026.02
24.8
2026.02
24.1
2026.02
23.69
2026.02
23.02