Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Wang, Dingzirui, Zhang, Xuanliang, Cao, Rongyu, Dou, Longxu, Luo, Xianzhen, Ma, Yingwei, Zhu, Qingfu, Che, Wanxiang, Li, Binhua, Huang, Fei, Li, Yongbin
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2506.23133
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866908427605770240
author Wang, Dingzirui
Zhang, Xuanliang
Cao, Rongyu
Dou, Longxu
Luo, Xianzhen
Ma, Yingwei
Zhu, Qingfu
Che, Wanxiang
Li, Binhua
Huang, Fei
Li, Yongbin
author_facet Wang, Dingzirui
Zhang, Xuanliang
Cao, Rongyu
Dou, Longxu
Luo, Xianzhen
Ma, Yingwei
Zhu, Qingfu
Che, Wanxiang
Li, Binhua
Huang, Fei
Li, Yongbin
contents Generating and voting multiple answers is an effective method to mitigate reasoning inconsistencies of large language models (LLMs). Prior works have shown that multiple reasoning formats outperform a single format when generating multiple answers. However, previous works using multiple formats rely on formats labeled by humans, which could be unsuitable for all tasks and have high labeling costs. To address this issue, we adapt suitable formats to the given tasks by generating and selecting formats. We first propose how to measure the reasoning error when generating multiple answers. Then, we introduce Format-Adapter, which utilizes LLMs to generate and select suitable reasoning formats by minimizing the error measurement we present. We conduct experiments on math and commonsense reasoning tasks, where Format-Adapter achieves a 4.3% performance improvement on average over previous works, demonstrating the effectiveness.
format Preprint
id arxiv_https___arxiv_org_abs_2506_23133
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Format-Adapter: Improving Reasoning Capability of LLMs by Adapting Suitable Format
Wang, Dingzirui
Zhang, Xuanliang
Cao, Rongyu
Dou, Longxu
Luo, Xianzhen
Ma, Yingwei
Zhu, Qingfu
Che, Wanxiang
Li, Binhua
Huang, Fei
Li, Yongbin
Computation and Language
Generating and voting multiple answers is an effective method to mitigate reasoning inconsistencies of large language models (LLMs). Prior works have shown that multiple reasoning formats outperform a single format when generating multiple answers. However, previous works using multiple formats rely on formats labeled by humans, which could be unsuitable for all tasks and have high labeling costs. To address this issue, we adapt suitable formats to the given tasks by generating and selecting formats. We first propose how to measure the reasoning error when generating multiple answers. Then, we introduce Format-Adapter, which utilizes LLMs to generate and select suitable reasoning formats by minimizing the error measurement we present. We conduct experiments on math and commonsense reasoning tasks, where Format-Adapter achieves a 4.3% performance improvement on average over previous works, demonstrating the effectiveness.
title Format-Adapter: Improving Reasoning Capability of LLMs by Adapting Suitable Format
topic Computation and Language
url https://arxiv.org/abs/2506.23133