Zero-shot Question Answering on Commonsense Reasoning Suite (PIQA, WinoGrande, HellaSwag, ARC)
82.7PIQA Accuracy (Zero-shot)
Dense
Updated 3d ago
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
2025.05 | 82.7 | 77.98 | 83.84 | 80.98 | 57.34 | 76.57 | |
2025.05 | 81.45 | 76.14 | 81.03 | 77.97 | 54.63 | 74.24 | |
2025.05 | 81.35 | 76.73 | 81.26 | 77.69 | 54.56 | 74.32 | |
2025.05 | 80.33 | 74.26 | 76.93 | 74.85 | 48.47 | 70.97 | |
2025.05 | 80.15 | 75.62 | 79.85 | 76.41 | 50.29 | 72.46 | |
2025.05 | 79.52 | 73.15 | 76.37 | 75.15 | 47.89 | 70.42 | |
2025.05 | 78.81 | 70.25 | 74.89 | 73.56 | 46.45 | 68.79 |