Policy-grounded Safety Evaluation on DynaBench Original
73.9Accuracy
LPG
Updated 15d ago
Evaluation Results
| Method | Links | ||
|---|---|---|---|
2026.05 | 73.9 | 61.25 | |
2026.05 | 71.82 | 69.94 | |
2026.05 | 67.4 | 53.29 | |
2026.05 | 58.56 | 31.19 | |
2026.05 | 56.9 | 23.02 |
LPG
| Method | Links | ||
|---|---|---|---|
2026.05 | 73.9 | 61.25 | |
2026.05 | 71.82 | 69.94 | |
2026.05 | 67.4 | 53.29 | |
2026.05 | 58.56 | 31.19 | |
2026.05 | 56.9 | 23.02 |