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

Commonsense Reasoning on WinoGrande (5-shot)

92.66Accuracy

ERNIE 5.0-Base

62.947270.661178.37586.0889Nov 14, 2025Nov 29, 2025Dec 15, 2025Dec 31, 2025Jan 15, 2026Jan 31, 2026Feb 16, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.02
92.66
2026.02
88.87
2026.02
87.61
2025.11
78.03
2025.11
77.95
2025.11
77.54
2026.02
77.43
2026.02
77.35
2026.02
77.35
2026.02
77.35
2026.02
77.35
2025.11
77.19
2025.11
77.06
2025.11
76.92
2025.11
76.66
2026.02
76.64
2025.11
76.56
2026.02
76.4
2025.11
76.4
2025.11
76.4
2025.11
76.38
2026.02
76.16
2026.02
76.09
2026.02
75.93
2025.11
75.35
2025.11
75.3
2025.11
74.98
2025.11
74.64
2025.11
74.59
2025.11
74.01
2025.11
73.69
2025.11
73.38
2025.11
73.17
2025.11
73.14
2025.11
72.96
2025.11
72.89
2025.11
72.69
2025.11
72.35
2025.11
72.22
2025.11
72.22
2025.11
72.05
2025.11
71.86
2025.11
71.82
2025.11
71.74
2025.11
71.19
2025.11
70.64
2025.11
70.56
2025.11
69.54
2026.02
69.53
2026.02
69.06
2026.02
68.98
2026.02
68.75
2026.02
68.51
2025.11
67.56
2025.11
67.29
2025.11
66.75
2025.11
66.54
2025.11
66.46
2025.11
66.17
2025.11
66.14
2025.11
66.12
2025.11
65.67
2025.11
64.38
2025.11
64.09