Toxicity Mitigation on RealToxicityPrompts challenging
6.2Avg Toxicity (Max)
RAD
Updated 4d ago
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
2026.01 | 6.2 | 5.5 | 35.79 | 0.5 | |
2026.01 | 7 | 5.8 | 19.46 | 0.6 | |
2026.01 | 7.2 | 6.1 | 47.23 | 2.3 | |
2026.01 | 7.8 | 6.6 | 30.02 | 1 | |
2026.01 | 8.4 | 6.8 | 22.4 | 2.6 | |
2026.01 | 9.8 | 7.8 | 17.03 | 2.2 | |
2026.01 | 11.6 | 8.7 | 8.72 | 6.2 | |
2026.01 | 12.9 | 9.9 | 7.06 | 5.4 | |
2026.01 | 13.4 | 10.2 | 7.21 | 4.6 | |
2026.01 | 14.4 | 12.2 | 4.2 | 0.3 | |
2026.01 | 14.6 | 11.2 | 5.33 | 6.7 | |
2026.01 | 15.6 | 12.2 | 5.53 | 0.3 | |
2026.01 | 15.9 | 12.2 | 6.25 | 6.6 | |
2026.01 | 16.7 | 12.9 | 6.07 | 6.9 | |
2026.01 | 17.4 | 13.4 | 5.18 | 8.8 | |
2026.01 | 18.4 | 16.4 | 5.21 | 6 | |
2026.01 | 19.4 | 15 | 5.19 | 9.8 | |
2026.01 | 20.9 | 15.7 | 5.5 | 12 | |
2026.01 | 24 | 18.6 | 4.47 | 14.1 | |
2026.01 | 27.5 | 22 | 4.41 | 20.2 | |
2026.01 | 28.1 | 22.6 | 5.67 | 21.5 | |
2026.01 | 29.7 | 24.5 | 7.46 | 25 | |
2026.01 | 31 | 24.7 | 4.67 | 26.1 | |
2026.01 | 33.1 | 27 | 4.26 | 28.7 | |
2026.01 | 33.1 | 26.7 | 4.3 | 28.9 | |
2026.01 | 33.3 | 27 | 3.82 | 29 | |
2026.01 | 33.4 | 27.1 | 6.63 | 32.1 | |
2026.01 | 34.1 | 27.8 | 4.28 | 29.5 | |
2026.01 | 34.2 | 27.7 | 4.28 | 30.4 | |
2026.01 | 41.9 | 33.6 | 4.39 | 40.1 | |
2026.01 | 42.5 | 34.1 | 3.65 | 40.4 | |
2026.01 | 52.7 | 46 | 4.18 | 56.2 | |
2026.01 | 54.6 | 46.2 | 6.42 | 57.2 | |
2026.01 | 55.7 | 45.5 | 4.17 | 59.6 | |
2026.01 | 57 | 47.5 | 3.51 | 61.5 | |
2026.01 | 57 | 46.8 | 4.18 | 61 | |
2026.01 | 57.2 | 47.5 | 3.54 | 61.7 | |
2026.01 | 57.5 | 47.6 | 4.27 | 63.1 | |
2026.01 | 57.6 | 47.8 | 3.57 | 62.9 | |
2026.01 | 57.9 | 47.9 | 3.53 | 63.4 | |
2026.01 | 57.9 | 48.1 | 4.2 | 62.1 | |
2026.01 | 58 | 48.2 | 4.18 | 62.4 | |
2026.01 | 58.1 | 48.9 | 4.53 | 62.4 | |
2026.01 | 58.7 | 48.8 | 4.2 | 63 | |
2026.01 | 59.1 | 49.5 | 4.13 | 63.9 | |
2026.01 | 59.2 | 49.2 | 4.2 | 63.8 |