Confidence Matters: Revisiting Intrinsic Self-Correction Capabilities of Large Language Models

TL;DR

This paper investigates the intrinsic self-correction abilities of Large Language Models, emphasizing the importance of their confidence levels, and introduces an IoE prompting framework to enhance their self-correction accuracy.

Contribution

It identifies confidence as a key factor in LLM self-correction and proposes a novel IoE prompting method to improve their intrinsic correction capabilities.

Findings

IoE prompting improves self-correction accuracy

LLMs can assess their own confidence levels

Framework demonstrates consistent performance gains

Abstract

The recent success of Large Language Models (LLMs) has catalyzed an increasing interest in their self-correction capabilities. This paper presents a comprehensive investigation into the intrinsic self-correction of LLMs, attempting to address the ongoing debate about its feasibility. Our research has identified an important latent factor - the "confidence" of LLMs - during the self-correction process. Overlooking this factor may cause the models to over-criticize themselves, resulting in unreliable conclusions regarding the efficacy of self-correction. We have experimentally observed that LLMs possess the capability to understand the "confidence" in their own responses. It motivates us to develop an "If-or-Else" (IoE) prompting framework, designed to guide LLMs in assessing their own "confidence", facilitating intrinsic self-corrections. We conduct extensive experiments and demonstrate that our IoE-based Prompt can achieve a consistent improvement regarding the accuracy of self-corrected responses over the initial answers. Our study not only sheds light on the underlying factors affecting self-correction in LLMs, but also introduces a practical framework that utilizes the IoE prompting principle to efficiently improve self-correction capabilities with "confidence". The code is available at https://github.com/MBZUAI-CLeaR/IoE-Prompting.git.