Toxicity Detection on Perturbed Text
85.57Performance (Insert)ContiGuard
Evaluation results
| Method | Links | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| ContiGuard2026.03 | 85.57 | 87.23 | 85.18 | 83.97 | 90.4 | 82.94 | 80.58 | 92.19 | 90.15 | 86.47 | 3 | |
| PespectiveAPI2026.03 | 77.2 | 64.61 | 70.48 | 60.51 | 64.84 | 63.29 | 50.93 | 66.38 | 64.37 | 64.73 | 25.9 | |
| DeTox2026.03 | 69.55 | 52.47 | 58.73 | 59.58 | 52.63 | 62.44 | 50.23 | 53.71 | 66.92 | 58.48 | 28.3 | |
| OpenAI Moderation2026.03 | 66.15 | 52.24 | 61.13 | 56.49 | 50.78 | 59.35 | 51 | 84.47 | 81.14 | 62.53 | 25.2 | |
| DetoxifyOriginal2026.03 | 61.59 | 63.76 | 61.9 | 57.73 | 56.96 | 59.27 | 56.03 | 71.87 | 75.66 | 62.75 | 29.8 | |
| ToxBERT2026.03 | 60.9 | 60.97 | 58.81 | 58.34 | 53.55 | 55.87 | 58.35 | 73.26 | 75.12 | 61.69 | 14.4 | |
| DetoxifyMultilingual2026.03 | 60.2 | 58.5 | 56.41 | 56.96 | 82.53 | 59.2 | 52.63 | 66.62 | 81.92 | 63.88 | 26.6 | |
| DetoxifyUnbiased2026.03 | 51.31 | 51 | 53.32 | 53.4 | 50.15 | 51.93 | 51.55 | 78.59 | 80.14 | 57.93 | 32.5 | |
| Paradetox2026.03 | 51.31 | 51.47 | 51.85 | 51.62 | 49.92 | 52.63 | 51.16 | 69.55 | 72.57 | 55.79 | 35.8 | |
| ToxRoBERTa2026.03 | 50.93 | 50.39 | 51.24 | 52.32 | 50.15 | 50.31 | 51.39 | 55.26 | 79.83 | 54.65 | 34 |