Evidence Retrieval on FEVER frozen 100-client instance (held-out)
55Accuracy
FedMark-FM
Updated 16d ago
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
2026.07 | 55 | 0.126 | 100 | 0 | |
2026.07 | 51.7 | 0.116 | 100 | 0 | |
2026.07 | 51.7 | 0.117 | 100 | 0.33 | |
2026.07 | 44.2 | 0.1 | 67 | 0.33 | |
2026.07 | 10.8 | 0.025 | 33 | 10.67 | |
2026.07 | 9.2 | 0.02 | 0 | 12 | |
2026.07 | 7.5 | 0.017 | 0 | 4.67 |