Fact Verification on FEVER (AUROC)
0.737AUROCVERDI LR
Evaluation Results
| Method | Links | |
|---|---|---|
| VERDI LRModel=GPT-4.1-mini2026.05 | 0.737 | |
| VERDI LRModel=Qwen3.5-4B2026.05 | 0.699 | |
| VERDI LRModel=GPT-5.4-mini2026.05 | 0.662 | |
| VERDI LRModel=Qwen3.5-9B2026.05 | 0.648 | |
| VERDI LRModel=Qwen3.5-27B2026.05 | 0.56 | |
| LogprobModel=Qwen3.5-4B2026.05 | 0.494 | |
| LogprobModel=Qwen3.5-27B2026.05 | 0.479 | |
| LogprobModel=Qwen3.5-9B2026.05 | 0.421 |