Group Robustness on CivilComments-WILDS (test)
67WG Accuracy
IRM
Updated 4d ago
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
2026.02 | 67 | 90 | - | |
2026.02 | 66.8 | 90.1 | - | |
2026.02 | 66.7 | 89.5 | - | |
2026.02 | 66.5 | 90.6 | - | |
2026.02 | 66.1 | 90.2 | - | |
2026.02 | 65.2 | 89.5 | - | |
2026.02 | 64.4 | 90.3 | - | |
2026.02 | 64.1 | 89.2 | - | |
2026.02 | 64 | 90.1 | - | |
2026.02 | 63.6 | 89.3 | - | |
2026.02 | 63.4 | 90.7 | - | |
2026.02 | 63.3 | 89.4 | - | |
2026.02 | 63.1 | 90.6 | - | |
2026.02 | 62.9 | 89.3 | - | |
2026.02 | 62.8 | 90.4 | - | |
2026.02 | 62.1 | 90.4 | - | |
2026.02 | 61.9 | 90.7 | - | |
2026.02 | 61.7 | 91.3 | - | |
2026.02 | 61.5 | 90 | - | |
2026.02 | 59.4 | 91.6 | - | |
2026.02 | 59.2 | 91.7 | - | |
2026.02 | 57.8 | 91.9 | - | |
2026.02 | 57.7 | 91.9 | - | |
2026.02 | 57.6 | 91.9 | - | |
2026.02 | 57.1 | 91.9 | - | |
2026.02 | 56.2 | 92 | - | |
2026.02 | 55.8 | 92.1 | - | |
2022.07 | 50.1 | 54.2 | 1.1 | |
2022.07 | 48.2 | 52.1 | 3.9 | |
2026.02 | 48 | 92.3 | - | |
2026.02 | 48 | 92.3 | - | |
2026.02 | 48 | 92.3 | - | |
2022.07 | 47.4 | 51.9 | 4.5 | |
2022.07 | 46.7 | 51.2 | 4.5 | |
2022.07 | 46.7 | 51.2 | 2.8 | |
2026.02 | 38.4 | 92 | - | |
2026.02 | 38.2 | 92 | - | |
2026.02 | 38.1 | 92 | - | |
2022.07 | 32.1 | 37.7 | 5.6 | |
2022.07 | 16 | 74.8 | 58.8 |