A six-stage text fact-checker that retrieves evidence from the supplied knowledge base and won the 2024 AVeriTeC shared task.
01 / Abstract
InFact decomposes text-claim verification into six stages and retrieves evidence from the supplied static AVeriTeC knowledge base. With GPT-4o as its backbone, it achieves an AVeriTeC score of 63% on the 2024 shared task’s test set, outperforming the other 20 participating teams. Its qualitative analysis identifies cases where the retrieved evidence supports a different conclusion from the benchmark annotation.
02 / Contributions
Ranks first among 21 systems in the 2024 AVeriTeC shared task with a 63% test-set score.
Turns claim verification into six explicit stages, including evidence retrieval from the static AVeriTeC knowledge base.
Examines cases where the retrieved evidence supports a different conclusion from the benchmark annotation.
03 / Citation
Rothermel, M., Braun, T., Rohrbach, M., & Rohrbach, A. (2024). InFact: A Strong Baseline for Automated Fact-Checking. Proceedings of the Seventh Fact Extraction and VERification Workshop (FEVER), 108–112.
@inproceedings{rothermel2024infact,
title = {{InFact}: A Strong Baseline for Automated Fact-Checking},
author = {Mark Rothermel and Tobias Braun and Marcus Rohrbach and Anna Rohrbach},
booktitle = {Proceedings of the Seventh Fact Extraction and VERification Workshop (FEVER)},
pages = {108--112},
address = {Miami, Florida, USA},
publisher = {Association for Computational Linguistics},
year = {2024},
doi = {10.18653/v1/2024.fever-1.12},
url = {https://aclanthology.org/2024.fever-1.12/}
}