Fact Verification On Fever
Métriques
Accuracy
FEVER
Résultats
Résultats de performance de divers modèles sur ce benchmark
Nom du modèle | Accuracy | FEVER | Paper Title | Repository |
|---|---|---|---|---|
| KGAT | 74.1 | 70.4 | Fine-grained Fact Verification with Kernel Graph Attention Network | |
| RoBERTa-Base Joint MSPP | 74.39 | - | Paragraph-based Transformer Pre-training for Multi-Sentence Inference | |
| DREAM | 76.85 | 70.60 | Reasoning Over Semantic-Level Graph for Fact Checking | - |
| RAG | 72.5 | - | Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks | |
| ProoFVer-SB | 79.47 | 76.82 | ProoFVer: Natural Logic Theorem Proving for Fact Verification | |
| RoBERTa-Base Joint MSPP Flexible | 75.36 | - | Paragraph-based Transformer Pre-training for Multi-Sentence Inference | |
| GEAR | 71.6 | 67.1 | GEAR: Graph-based Evidence Aggregating and Reasoning for Fact Verification |
0 of 7 row(s) selected.