| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| Bot-Adversarial Dialogue (BAD) 1.0 (test) | CLICK | Toxicity Probability0.084 | 10 | 4d ago | |
| BAD (test) | CLICK | Toxicity Reduction37 | 10 | 4d ago | |
| BAD (val) | CLICK | Toxicity Proportion11 | 7 | 4d ago | |
| BEEP (test) | Overall Offensiveness1.468 | 5 | 4d ago | ||
| KOLD (test) | K/DA | Overall Offensiveness Score1.606 | 5 | 4d ago | |
| Ours (test) | K/DA | Overall Offensiveness Score1.145 | 5 | 4d ago |