Sentence-level binary bias detection on BABE (test)
85.3F1 ScoreHierBias
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| HierBias2026.04 | 85.3 | 0.723 | 84.9 | 85.8 | |
| bias-detector2026.04 | 83.1 | 0.693 | 82.6 | 83.7 | |
| LLM prompting2026.04 | 81.9 | 0.682 | 81.5 | 82.3 | |
| DA-RoBERTa2026.04 | 81.4 | 0.674 | 80.9 | 82 | |
| RoBERTa-base2026.04 | 76.3 | 0.612 | 75.8 | 76.9 | |
| BERT-base2026.04 | 74.1 | 0.583 | 73.5 | 74.8 | |
| LR + TF-IDF2026.04 | 62.3 | 0.412 | 61.8 | 62.9 |