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

Zero-shot Reasoning on WinoGrande

70Accuracy

Baseline

-1.7616.8735.554.13Jun 15, 2024Oct 2, 2024Jan 19, 2025May 9, 2025Aug 26, 2025Dec 13, 2025Apr 2, 2026
Updated 9d ago

Evaluation Results

MethodLinks
2026.04
70
2026.02
69
2024.06
68.43
2026.04
68
2026.04
68
2026.02
67.6
2026.04
64
2026.02
63.5
2024.06
63.22
2026.04
62
2026.04
62
2026.04
62
2024.06
61.6
2026.04
61
2024.06
60.93
2024.06
60.38
2024.06
60.3
2025.10
59.48
2025.10
59.27
2026.02
58.5
2025.10
58.4
2026.04
58
2025.10
57.77
2025.10
57.45
2024.06
56.27
2024.06
55.72
2025.10
55.56
2024.06
55.41
2025.10
55.24
2026.04
55
2026.02
54.4
2025.10
54.14
2024.06
54.06
2025.10
53.51
2026.02
53.4
2025.10
53.35
2024.06
53.12
2026.04
53
2025.10
52.95
2024.06
52.8
2025.10
52.8
2024.06
52.64
2024.06
52.17
2026.04
52
2024.06
51.7
2026.02
50.5
2024.06
50.11
2026.04
48
2026.04
17
2026.04
9
2026.04
6
2026.04
5
2026.04
2
2026.04
1