FactCG: Enhancing Fact Checkers with Graph-Based Multi-Hop Data

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Lei, Deren, Li, Yaxi, Li, Siyao, Hu, Mengya, Xu, Rui, Archer, Ken, Wang, Mingyu, Ching, Emily, Deng, Alex
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866908203317460992
author Lei, Deren
Li, Yaxi
Li, Siyao
Hu, Mengya
Xu, Rui
Archer, Ken
Wang, Mingyu
Ching, Emily
Deng, Alex
author_facet Lei, Deren
Li, Yaxi
Li, Siyao
Hu, Mengya
Xu, Rui
Archer, Ken
Wang, Mingyu
Ching, Emily
Deng, Alex
contents Prior research on training grounded factuality classification models to detect hallucinations in large language models (LLMs) has relied on public natural language inference (NLI) data and synthetic data. However, conventional NLI datasets are not well-suited for document-level reasoning, which is critical for detecting LLM hallucinations. Recent approaches to document-level synthetic data generation involve iteratively removing sentences from documents and annotating factuality using LLM-based prompts. While effective, this method is computationally expensive for long documents and limited by the LLM's capabilities. In this work, we analyze the differences between existing synthetic training data used in state-of-the-art models and real LLM output claims. Based on our findings, we propose a novel approach for synthetic data generation, CG2C, that leverages multi-hop reasoning on context graphs extracted from documents. Our fact checker model, FactCG, demonstrates improved performance with more connected reasoning, using the same backbone models. Experiments show it even outperforms GPT-4-o on the LLM-Aggrefact benchmark with much smaller model size.
format Preprint
id arxiv_https___arxiv_org_abs_2501_17144
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle FactCG: Enhancing Fact Checkers with Graph-Based Multi-Hop Data
Lei, Deren
Li, Yaxi
Li, Siyao
Hu, Mengya
Xu, Rui
Archer, Ken
Wang, Mingyu
Ching, Emily
Deng, Alex
Computation and Language
Artificial Intelligence
Prior research on training grounded factuality classification models to detect hallucinations in large language models (LLMs) has relied on public natural language inference (NLI) data and synthetic data. However, conventional NLI datasets are not well-suited for document-level reasoning, which is critical for detecting LLM hallucinations. Recent approaches to document-level synthetic data generation involve iteratively removing sentences from documents and annotating factuality using LLM-based prompts. While effective, this method is computationally expensive for long documents and limited by the LLM's capabilities. In this work, we analyze the differences between existing synthetic training data used in state-of-the-art models and real LLM output claims. Based on our findings, we propose a novel approach for synthetic data generation, CG2C, that leverages multi-hop reasoning on context graphs extracted from documents. Our fact checker model, FactCG, demonstrates improved performance with more connected reasoning, using the same backbone models. Experiments show it even outperforms GPT-4-o on the LLM-Aggrefact benchmark with much smaller model size.
title FactCG: Enhancing Fact Checkers with Graph-Based Multi-Hop Data
topic Computation and Language
Artificial Intelligence
url https://arxiv.org/abs/2501.17144