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

Multi-task Language Understanding on MMLU (Inference Latency)

2.07Average Inference Time (s)

CoT

-2.08225.94453.9781.996Sep 24, 2025
Updated 24d ago

Evaluation Results

MethodLinks
2025.09
2.07
2025.09
2.29
2025.09
2.73
2025.09
3.48
2025.09
3.9
2025.09
4.28
2025.09
4.93
2025.09
5.42
2025.09
6.86
2025.09
7.61
2025.09
10.67
2025.09
11.38
2025.09
14.44
2025.09
14.55
2025.09
16.15
2025.09
17.39
2025.09
20.38
2025.09
23.89
2025.09
28.14
2025.09
29.6
2025.09
29.82
2025.09
36.49
2025.09
41.58
2025.09
42.96
2025.09
47.76
2025.09
51.78
2025.09
55.91
2025.09
64.03
2025.09
77.31
2025.09
105.87