Fact-checking on FEVEROUS
89.4F1 MacroNone (Original Claims)
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| None (Original Claims)Input Setting=Gold-Evidence2026.01 | 89.4 | 96.4 | 89.4 | — | |
| Word SwapInput Setting=Gold-Evidence2026.01 | 88.3 | 95.5 | 88.4 | -0.01 | |
| ParaphrasingInput Setting=Gold-Evidence2026.01 | 88.1 | 95.4 | 88.2 | -0.012 | |
| Character PerturbationsInput Setting=Gold-Evidence2026.01 | 87.7 | 95.3 | 87.8 | -0.016 | |
| Synonym SubstitutionInput Setting=Gold-Evidence2026.01 | 86.3 | 94.8 | 86.3 | -0.031 | |
| Persuasion (Blind)Input Setting=Gold-Evidence2026.01 | 72.2 | 87 | 72.6 | -0.168 | |
| None (Original Claims)Input Setting=Claim-Only2026.01 | 62.7 | 68.6 | 64.1 | — | |
| Character PerturbationsInput Setting=Claim-Only2026.01 | 61.9 | 67.5 | 63.2 | -0.009 | |
| Synonym SubstitutionInput Setting=Claim-Only2026.01 | 61.8 | 66.6 | 62.3 | -0.018 | |
| ParaphrasingInput Setting=Claim-Only2026.01 | 61 | 66.9 | 62.9 | -0.012 | |
| Word SwapInput Setting=Claim-Only2026.01 | 60.8 | 66.2 | 60.9 | -0.032 | |
| Persuasion (Blind)Input Setting=Claim-Only2026.01 | 56.7 | 59.7 | 57.5 | -0.066 | |
| Persuasion (Oracle)Input Setting=Gold-Evidence2026.01 | 20.3 | 12.9 | 25 | -0.644 | |
| Persuasion (Oracle)Input Setting=Claim-Only2026.01 | 1 | 0.1 | 1 | -0.631 |