Format-Adapter: Improving Reasoning Capability of LLMs by Adapting Suitable Format

TL;DR

This paper introduces Format-Adapter, a method that automatically adapts reasoning formats for large language models to improve their reasoning accuracy without relying on human-labeled formats.

Contribution

The paper proposes a novel approach to automatically generate and select suitable reasoning formats for LLMs, reducing labeling costs and enhancing reasoning performance.

Findings

Achieves 4.3% average performance improvement over previous methods

Effectively measures reasoning errors to select optimal formats

Demonstrates effectiveness on math and commonsense reasoning tasks

Abstract

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.