HateXplain
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Hate speech detection | HateXplain (test) | Macro F1 Score85.13 | 24 | |
| Hate speech classification and explainability | HateXplain (test) | IOU F10.3176 | 22 | |
| Classification Probing | HateXplain (test) | Probe Accuracy (Best Layer)79.1 | 21 | |
| Toxicity Detection | HateXplain | AUC90.44 | 21 | |
| Hate Speech Detection | HateXplain (held-out) | F1 Score47.3 | 14 | |
| Hate Speech Classification | HateXplain | Accuracy0.724 | 7 | |
| Rationalization | HateXplain Synthetic Skew (test) | HI-F10.4281 | 7 | |
| Classification | HateXplain Synthetic Skew (test) | Clf-F173.15 | 7 | |
| Hate Speech Detection | HateXplain Muslim-focused (test) | Accuracy75.1 | 7 | |
| Text Classification | HateXplain (val) | Accuracy35.2 | 6 | |
| Rationalization | HateXplain (test) | HI-F142.62 | 5 | |
| Hate Speech Classification | HateXplain standard (test) | Accuracy68.6 | 5 | |
| Concept vector stability | HateXplain | Mean Abs-Cosine Similarity0.98 | 3 | |
| Text Classification | HateXplain (test) | Accuracy34.3 | 3 |
Showing 14 of 14 rows