Commonsense Reasoning on WinoGAViL
65.2Accuracy (5-6)
ReasonSigLIP2
Updated 29d ago
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
2026.06 | 65.2 | 60.1 | 64 | |
2026.06 | 64.4 | 61.5 | 63.7 | |
2026.06 | 63.7 | 56.5 | 62 | |
2026.06 | 63.2 | 58.4 | 62.1 | |
2026.06 | 63 | 57.7 | 61.7 | |
2026.06 | 62.5 | 56.6 | 61.1 | |
2026.06 | 62.3 | 54.6 | 60.2 | |
2026.06 | 62.1 | 56.8 | 60.8 | |
2026.06 | 62.1 | 60.1 | 61.6 | |
2026.06 | 59.6 | 53.6 | 58.2 | |
2026.06 | 58.7 | 50.9 | 56.8 | |
2026.06 | 58.3 | 51 | 56.5 | |
2026.06 | 57.8 | 49.6 | 55.8 | |
2026.06 | 56.8 | 50.5 | 55.3 | |
2026.06 | 56.5 | 50.5 | 55 | |
2026.06 | 55.6 | 45.8 | 53.3 | |
2026.06 | 54.4 | 46.8 | 53.7 | |
2026.06 | 53.8 | 44 | 51.5 | |
2026.06 | 53.5 | 41.5 | 50.6 | |
2026.06 | 53.3 | 41.1 | 50.4 | |
2026.06 | 53.2 | 42.6 | 50.6 | |
2026.06 | 53 | 41.4 | 50.2 |