Research map / Large language models / Language, safety and interpretability
FEVEROUS: Fact Extraction and VERification Over Unstructured and Structured information
Rami Aly, Zhijiang Guo, Michael Sejr Schlichtkrull, James Thorne, Andreas Vlachos, Christos Christodoulopoulos, Oana Cocarascu, Arpit Mittal
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Combine sentence and table-cell evidence instead of using sentences only | FEVEROUS baseline · FEVEROUS test | Score: 0.12 -> 0.18 | Improved |
| Combine sentence and table-cell evidence instead of using table cells only | FEVEROUS baseline · FEVEROUS test | Score: 0.05 -> 0.18 | Improved |
| Combine sentence and table-cell evidence instead of using sentences only | FEVEROUS baseline · FEVEROUS test | Evidence: 0.19 -> 0.29 | Improved |
| Combine sentence and table-cell evidence instead of using table cells only | FEVEROUS baseline · FEVEROUS test | Evidence: 0.07 -> 0.29 | Improved |
| Add NLI pre-training (ablation) | RoBERTa verdict predictor with gold evidence · FEVEROUS dev | Overall macro-averaged F1: 0.53 -> 0.62 | Improved |
| Add NLI pre-training (ablation) | RoBERTa verdict predictor with gold evidence · FEVEROUS dev | NEI F1: 0.05 -> 0.09 | Improved |
| Add NEI sampling to the NLI-pre-trained model (ablation) | RoBERTa verdict predictor with gold evidence · FEVEROUS dev | Overall macro-averaged F1: 0.62 -> 0.70 | Improved |
| Add NEI sampling to the NLI-pre-trained model (ablation) | RoBERTa verdict predictor with gold evidence · FEVEROUS dev | NEI F1: 0.09 -> 0.34 | Improved |
| Add entity matching to TF-IDF retrieval (ablation) | FEVEROUS retriever · FEVEROUS | Document coverage: 49% -> 69% | Improved |
Closest papers by meaning
- Unsupervised Pretraining for Fact Verification by Language Model Distillation (ICLR 2024 · similarity 0.87)
- Heterogeneous Graph Reasoning for Fact Checking over Texts and Tables (AAAI 2024 · similarity 0.85)
- CFEVER: A Chinese Fact Extraction and VERification Dataset (AAAI 2024 · similarity 0.85)
- Beyond Text: Fine-Grained Multi-Modal Fact Verification with Hypergraph Transformers (AAAI 2025 · similarity 0.85)
- AVerImaTeC: A Dataset for Automatic Verification of Image-Text Claims with Evidence from the Web (NeurIPS 2025 · similarity 0.85)
- DEFAME: Dynamic Evidence-based FAct-checking with Multimodal Experts (ICML 2025 · similarity 0.84)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
