Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

Commonsense Reasoning on HellaSwag (Zero-shot)

60.04Zero-shot Accuracy

Dense

24.347233.613642.8852.1464Oct 2, 2025
Updated 2d ago

Evaluation Results

MethodLinks
2025.10
60.04
2025.10
57.45
2025.10
56.96
2025.10
56.71
2025.10
56.59
2025.10
56.51
2025.10
55.95
2025.10
54.39
2025.10
54.22
2025.10
52.09
2025.10
51.97
2025.10
51.49
2025.10
49.38
2025.10
48.71
2025.10
47.52
2025.10
44.66
2025.10
41.62
2025.10
40.86
2025.10
40.05
2025.10
38.03
2025.10
36.27
2025.10
30.65
2025.10
29.67
2025.10
29.25
2025.10
28.65
2025.10
27.77
2025.10
27.53
2025.10
27.44
2025.10
27.37
2025.10
27.08
2025.10
26.81
2025.10
26.02
2025.10
25.88
2025.10
25.87
2025.10
25.8
2025.10
25.72