Enregistré dans:
Détails bibliographiques
Auteurs principaux: Ryu, Hyun, Chu, Gyouk, Betz, Gregor, Yang, Eunho, Rose, Carolyn, Welleck, Sean
Format: Preprint
Publié: 2026
Sujets:
Accès en ligne:https://arxiv.org/abs/2603.17432
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866910220518686720
author Ryu, Hyun
Chu, Gyouk
Betz, Gregor
Yang, Eunho
Rose, Carolyn
Welleck, Sean
author_facet Ryu, Hyun
Chu, Gyouk
Betz, Gregor
Yang, Eunho
Rose, Carolyn
Welleck, Sean
contents To think critically about arguments, human learners are trained to identify, reconstruct, and evaluate arguments. Argument reconstruction is especially important because it makes an argument's underlying inferences explicit. However, it remains unclear whether LLMs can similarly enhance their critical thinking ability by learning to reconstruct arguments. To address this question, we introduce a holistic framework with three contributions. We (1) propose an engine that automatically reconstructs arbitrary arguments (GAAR), (2) synthesize a new high-quality argument reconstruction dataset (Arguinas) using the GAAR engine, and (3) investigate whether learning argument reconstruction benefits downstream critical thinking tasks. Our experimental results show that, across seven critical thinking tasks, models trained to learn argument reconstruction outperform models that do not, with the largest performance gains observed when training on the proposed Arguinas dataset.
format Preprint
id arxiv_https___arxiv_org_abs_2603_17432
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Argument Reconstruction as Supervision for Critical Thinking in LLMs
Ryu, Hyun
Chu, Gyouk
Betz, Gregor
Yang, Eunho
Rose, Carolyn
Welleck, Sean
Computation and Language
To think critically about arguments, human learners are trained to identify, reconstruct, and evaluate arguments. Argument reconstruction is especially important because it makes an argument's underlying inferences explicit. However, it remains unclear whether LLMs can similarly enhance their critical thinking ability by learning to reconstruct arguments. To address this question, we introduce a holistic framework with three contributions. We (1) propose an engine that automatically reconstructs arbitrary arguments (GAAR), (2) synthesize a new high-quality argument reconstruction dataset (Arguinas) using the GAAR engine, and (3) investigate whether learning argument reconstruction benefits downstream critical thinking tasks. Our experimental results show that, across seven critical thinking tasks, models trained to learn argument reconstruction outperform models that do not, with the largest performance gains observed when training on the proposed Arguinas dataset.
title Argument Reconstruction as Supervision for Critical Thinking in LLMs
topic Computation and Language
url https://arxiv.org/abs/2603.17432