Bias Detection on BABE (test)
80.8Macro F1mini_listwise_BT
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| mini_listwise_BTModel=mini, Strategy=Listwise comparison, Rating Algorithm=Bradley-Terry (BT), Rounds=32025.12 | 80.8 | 84.4 | 80.3 | 77.6 | |
| BERT + distantModel=BERT, Strategy=Distant supervision2025.12 | 80.4 | — | — | — | |
| mini_directModel=mini, Strategy=Direct assessment2025.12 | 79.2 | 85.1 | 78 | 74.1 | |
| nano_24_pairwise_BTModel=nano, Strategy=Pairwise comparison, Rating Algorithm=Bradley-Terry (BT), Rounds=242025.12 | 79 | 77.6 | 79.6 | 80.3 | |
| nano_24_pairwise_EloModel=nano, Strategy=Pairwise comparison, Rating Algorithm=Elo, Rounds=242025.12 | 78.8 | 79.2 | 79 | 78.4 | |
| nano_listwise_BTModel=nano, Strategy=Listwise comparison, Rating Algorithm=Bradley-Terry (BT), Rounds=32025.12 | 78.4 | 79.6 | 78.4 | 77.3 | |
| mini_listwise_EloModel=mini, Strategy=Listwise comparison, Rating Algorithm=Elo, Rounds=32025.12 | 77.8 | 78.3 | 77.9 | 77.2 | |
| nano_listwise_EloModel=nano, Strategy=Listwise comparison, Rating Algorithm=Elo, Rounds=32025.12 | 76.1 | 76.1 | 76.3 | 75.6 |