Fact-checking on FEVEROUS-S
72.55Macro F1RRC
Evaluation Results
| Method | Links | |
|---|---|---|
| RRCMethod Category=LLM-based reasoning methods, Backbone Model=DeepSeek-Coder, k=5, threshold=0.852026.01 | 72.55 | |
| BiDevMethod Category=LLM-based reasoning methods, Source=Reproduction, Backbone Model=DeepSeek-Coder, k=5, threshold=0.852026.01 | 68.61 | |
| BootstrappingMethod Category=LLM-based reasoning methods, Source=Reproduction, Backbone Model=DeepSeek-Coder, k=5, threshold=0.852026.01 | 66.12 | |
| ProgramFCMethod Category=LLM-based reasoning methods, Source=Reproduction, Backbone Model=DeepSeek-Coder, k=5, threshold=0.852026.01 | 65.59 | |
| FOLKMethod Category=LLM-based reasoning methods, Source=Reproduction, Backbone Model=DeepSeek-Coder, k=5, threshold=0.852026.01 | 64.07 | |
| FLAN-T5Method Category=LLM-based ICL methods, Source=Reported in [9], k=5, threshold=0.852026.01 | 63.73 | |
| CodexMethod Category=LLM-based ICL methods, Source=Reported in [9], k=5, threshold=0.852026.01 | 62.58 | |
| DeBERTaV3-NLIMethod Category=Finetuned methods, Source=Reported in [9], k=5, threshold=0.852026.01 | 58.81 | |
| RoBERTa-NLIMethod Category=Finetuned methods, Source=Reported in [9], k=5, threshold=0.852026.01 | 57.8 | |
| MULTIVERSMethod Category=Finetuned methods, Source=Reported in [9], k=5, threshold=0.852026.01 | 56.61 | |
| LisT5Method Category=Pretrained methods, Source=Reported in [9], k=5, threshold=0.852026.01 | 54.15 | |
| BERT-FCMethod Category=Pretrained methods, Source=Reported in [9], k=5, threshold=0.852026.01 | 51.67 |