Automated Fact-Checking on PubHealth
69.91AccuracyCREDENCE-decomp
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| CREDENCE-decompSetting=CREDENCE-decomp, Decomposition model=Qwen3-8B, NLI verifier=Qwen NLI verifier2026.06 | 69.91 | 0.0111 | 0.0132 | |
| No decompSetting=No decomp, NLI verifier=Qwen NLI verifier2026.06 | 68.8 | — | — | |
| FactScore-decompSetting=FactScore-decomp, Decomposition model=Qwen3-8B, NLI verifier=Qwen NLI verifier2026.06 | 68.59 | 0.0021 | — |