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

Commonsense Reasoning on HellaSwag (Accuracy)

84HellaSwag Accuracy

LLaMA-2 70B

68.19272.29676.480.504Jun 16, 2025Aug 5, 2025Sep 24, 2025Nov 13, 2025Jan 2, 2026Feb 21, 2026Apr 12, 2026
Updated 4d ago

Evaluation Results

MethodLinks
2026.03
84
2026.03
84
2026.03
82.7
2026.03
81.2
2026.03
81.2
2026.03
81.02
2026.03
80.95
2026.03
80.5
2026.03
80.5
2026.03
80.5
2026.03
80.47
2026.03
80.38
2026.03
80.26
2026.03
79.9
2026.03
79.9
2026.03
79.63
2026.03
79.6
2026.03
79.6
2026.03
79.51
2026.03
79.28
2026.04
79.21
2026.03
79.05
2026.03
78.67
2026.03
78.52
2026.03
78.3
2026.03
78.3
2026.04
78.21
2026.04
78
2026.03
77.74
2026.03
77.6
2026.03
77.6
2026.04
77.6
2026.03
77.57
2026.03
77.5
2026.04
77.5
2026.03
77.4
2026.03
77.4
2026.03
77.3
2026.03
77.3
2026.04
77.3
2026.03
77.27
2026.03
77.02
2026.03
77.02
2026.04
77
2026.04
76.7
2026.04
76.4
2026.04
76.3
2026.04
76.23
2026.04
75.9
2026.04
75.54
2026.03
75.5
2026.04
75.4
2026.03
75.29
2026.03
75.07
2026.04
75
2026.04
75
2026.04
74.9
2026.04
74.9
2026.04
74.9
2026.04
74.9
2026.04
74.9
2026.04
74.61
2026.04
74.5
2026.04
74.5
2026.04
74.4
2026.04
74.37
2026.03
74.3
2026.03
74.3
2026.03
74.23
2026.03
74.2
2026.03
74.2
2026.03
74.2
2026.03
74.15
2026.04
74.1
2026.03
74
2026.03
73.75
2026.03
73.6
2026.03
73.6
2026.04
73.3
2026.04
73.23
2026.03
73
2026.04
72.96
2026.04
72.8
2026.04
72.8
2026.03
72.7
2026.03
72.6
2026.04
72.5
2026.04
72.4
2026.04
72.09
2026.04
72
2026.03
71.7
2026.03
71.7
2026.03
71.6
2026.03
71.55
2026.04
71.5
2026.03
71.17
2026.04
70.38
2026.04
69.2
2025.06
68.9
2026.04
68.8
Showing 100 of 350 rows