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

Language Understanding on MMLU (Delta Accuracy)

37.28Delta Accuracy

LMNet

-1.24168.759218.7628.7608May 19, 2025
Updated 2mo ago

Evaluation Results

MethodLinks
2025.05
37.28
2025.05
35.83
2025.05
35.16
2025.05
35.02
2025.05
34.75
2025.05
33.85
2025.05
27.57
2025.05
27.19
2025.05
24.83
2025.05
24.65
2025.05
24.3
2025.05
23.35
2025.05
22.5
2025.05
22.36
2025.05
21.5
2025.05
21.3
2025.05
20.93
2025.05
20.69
2025.05
19.48
2025.05
19.22
2025.05
18.54
2025.05
13.1
2025.05
11.69
2025.05
9.88
2025.05
7.4
2025.05
6.05
2025.05
5.01
2025.05
3.63
2025.05
1.69
2025.05
0.24