arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2310.01798cs.CLcs.AI
收藏

大型语言模型尚无法自我纠正推理

Large Language Models Cannot Self-Correct Reasoning Yet

University of Illinois at Urbana-Champaign(伊利诺伊大学厄巴纳-香槟分校);Google DeepMind(谷歌DeepMind)

Jie Huang, Xinyun Chen, Swaroop Mishra, Huaixiu Steven Zheng, Adams Wei Yu, Xinying Song, Denny Zhou

AI总结:

本文批判性检验大语言模型的内在自我纠正能力,发现无外部反馈时模型难以自我纠正推理回答,甚至性能下降,并据此提出未来研究与应用建议。

摘要:

大型语言模型(LLMs)作为一种突破性技术,凭借其在各种应用中无与伦比的文本生成能力而崭露头角。然而,人们对其生成内容的准确性和适当性仍存担忧。一种当代方法论——自我纠正,已被提出作为这些问题的补救措施。基于这一前提,本文批判性地审视了自我纠正在大语言模型中的作用和有效性,揭示了其真正的潜力和局限性。我们调查的核心是内在自我纠正的概念,即大语言模型仅凭其固有能力,在没有外部反馈辅助的情况下,尝试纠正其初始回答。在推理的背景下,我们的研究表明,大语言模型在没有外部反馈的情况下难以自我纠正其回答,有时甚至在进行自我纠正后性能反而下降。基于这些见解,我们为该领域的未来研究和实际应用提供了建议。

英文摘要:

Large Language Models (LLMs) have emerged as a groundbreaking technology with their unparalleled text generation capabilities across various applications. Nevertheless, concerns persist regarding the accuracy and appropriateness of their generated content. A contemporary methodology, self-correction, has been proposed as a remedy to these issues. Building upon this premise, this paper critically examines the role and efficacy of self-correction within LLMs, shedding light on its true potential and limitations. Central to our investigation is the notion of intrinsic self-correction, whereby an LLM attempts to correct its initial responses based solely on its inherent capabilities, without the crutch of external feedback. In the context of reasoning, our research indicates that LLMs struggle to self-correct their responses without external feedback, and at times, their performance even degrades after self-correction. Drawing from these insights, we offer suggestions for future research and practical applications in this field.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑