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

Commonsense Reasoning on WinoGrande (WG)

77.5Accuracy

Qwen3-8B

49.596856.840964.08571.3291Jun 11, 2026Jun 14, 2026Jun 18, 2026Jun 21, 2026Jun 25, 2026Jun 28, 2026Jul 2, 2026
Updated 22d ago

Evaluation Results

MethodLinks
2026.06
77.5
2026.06
77.35
2026.06
76.1
2026.06
75.9
2026.06
74.5
2026.06
74.5
2026.06
74.2
2026.06
73.5
2026.07
72.8
2026.07
72.3
2026.07
71.9
2026.07
71.3
2026.06
71.27
2026.07
70.9
2026.06
70.64
2026.07
70.6
2026.07
70.5
2026.06
68.75
2026.07
68.4
2026.06
68.35
2026.07
68.2
2026.07
66.4
2026.07
65.7
2026.06
65.19
2026.06
64.9
2026.06
64.8
2026.06
64.72
2026.06
63.1
2026.06
61.9
2026.07
56.1
2026.06
54.9
2026.06
54.9
2026.06
54.4
2026.06
51.7
2026.07
51.54
2026.07
51.38
2026.07
51.38
2026.07
50.67