Guardado en:
Detalles Bibliográficos
Autores principales: Dipta, Shubhashis Roy, Ferraro, Francis
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:https://arxiv.org/abs/2508.16838
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866912611165011968
author Dipta, Shubhashis Roy
Ferraro, Francis
author_facet Dipta, Shubhashis Roy
Ferraro, Francis
contents Prior work has shown that presupposition in generated questions can introduce unverified assumptions, leading to inconsistencies in claim verification. Additionally, prompt sensitivity remains a significant challenge for large language models (LLMs), resulting in performance variance as high as 3-6%. While recent advancements have reduced this gap, our study demonstrates that prompt sensitivity remains a persistent issue. To address this, we propose a structured and robust claim verification framework that reasons through presupposition-free, decomposed questions. Extensive experiments across multiple prompts, datasets, and LLMs reveal that even state-of-the-art models remain susceptible to prompt variance and presupposition. Our method consistently mitigates these issues, achieving up to a 2-5% improvement.
format Preprint
id arxiv_https___arxiv_org_abs_2508_16838
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle If We May De-Presuppose: Robustly Verifying Claims through Presupposition-Free Question Decomposition
Dipta, Shubhashis Roy
Ferraro, Francis
Computation and Language
Prior work has shown that presupposition in generated questions can introduce unverified assumptions, leading to inconsistencies in claim verification. Additionally, prompt sensitivity remains a significant challenge for large language models (LLMs), resulting in performance variance as high as 3-6%. While recent advancements have reduced this gap, our study demonstrates that prompt sensitivity remains a persistent issue. To address this, we propose a structured and robust claim verification framework that reasons through presupposition-free, decomposed questions. Extensive experiments across multiple prompts, datasets, and LLMs reveal that even state-of-the-art models remain susceptible to prompt variance and presupposition. Our method consistently mitigates these issues, achieving up to a 2-5% improvement.
title If We May De-Presuppose: Robustly Verifying Claims through Presupposition-Free Question Decomposition
topic Computation and Language
url https://arxiv.org/abs/2508.16838