A Closer Look at the Self-Verification Abilities of Large Language Models in Logical Reasoning

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Hong, Ruixin, Zhang, Hongming, Pang, Xinyu, Yu, Dong, Zhang, Changshui
Format: Preprint
Published: 2023
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916172841091072
author Hong, Ruixin
Zhang, Hongming
Pang, Xinyu
Yu, Dong
Zhang, Changshui
author_facet Hong, Ruixin
Zhang, Hongming
Pang, Xinyu
Yu, Dong
Zhang, Changshui
contents Logical reasoning has been an ongoing pursuit in the field of AI. Despite significant advancements made by large language models (LLMs), they still struggle with complex logical reasoning problems. To enhance reasoning performance, one promising direction is scalable oversight, which requires LLMs to identify their own errors and then improve by themselves. Various self-verification methods have been proposed in pursuit of this goal. Nevertheless, whether existing models understand their own errors well is still under investigation. In this paper, we take a closer look at the self-verification abilities of LLMs in the context of logical reasoning, focusing on their ability to identify logical fallacies accurately. We introduce a dataset, FALLACIES, containing 232 types of reasoning fallacies categorized in a hierarchical taxonomy. By conducting exhaustive experiments on FALLACIES, we obtain comprehensive and detailed analyses of a series of models on their verification abilities. Our main findings suggest that existing LLMs could struggle to identify fallacious reasoning steps accurately and may fall short of guaranteeing the validity of self-verification methods. Drawing from these observations, we offer suggestions for future research and practical applications of self-verification methods.
format Preprint
id arxiv_https___arxiv_org_abs_2311_07954
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle A Closer Look at the Self-Verification Abilities of Large Language Models in Logical Reasoning
Hong, Ruixin
Zhang, Hongming
Pang, Xinyu
Yu, Dong
Zhang, Changshui
Artificial Intelligence
Computation and Language
Logical reasoning has been an ongoing pursuit in the field of AI. Despite significant advancements made by large language models (LLMs), they still struggle with complex logical reasoning problems. To enhance reasoning performance, one promising direction is scalable oversight, which requires LLMs to identify their own errors and then improve by themselves. Various self-verification methods have been proposed in pursuit of this goal. Nevertheless, whether existing models understand their own errors well is still under investigation. In this paper, we take a closer look at the self-verification abilities of LLMs in the context of logical reasoning, focusing on their ability to identify logical fallacies accurately. We introduce a dataset, FALLACIES, containing 232 types of reasoning fallacies categorized in a hierarchical taxonomy. By conducting exhaustive experiments on FALLACIES, we obtain comprehensive and detailed analyses of a series of models on their verification abilities. Our main findings suggest that existing LLMs could struggle to identify fallacious reasoning steps accurately and may fall short of guaranteeing the validity of self-verification methods. Drawing from these observations, we offer suggestions for future research and practical applications of self-verification methods.
title A Closer Look at the Self-Verification Abilities of Large Language Models in Logical Reasoning
topic Artificial Intelligence
Computation and Language
url https://arxiv.org/abs/2311.07954