推理感知的自洽性：利用推理路径实现高效的大语言模型采样
Reasoning Aware Self-Consistency: Leveraging Reasoning Paths for Efficient LLM Sampling
AI总结：
本文提出推理感知的自洽性框架RASC，通过动态评估推理质量和答案一致性来指导早停与推理依据选择，在保持准确率的同时将采样量减少约70%，有效解决大语言模型推理中的效率与忠实度权衡问题。
摘要：
自洽性（Self-Consistency）通过采样多条推理路径来缓解大语言模型（LLMs）中的幻觉问题，但缺乏系统性的方法来确定最优采样数量或选择最忠实的推理依据。为了解决这一局限，我们提出了推理感知的自洽性（Reasoning-Aware Self-Consistency，RASC），这是一个新颖的框架，通过动态评估输出和推理依据来提升采样效率和推理忠实度。RASC 评估每个生成样本的推理质量和答案一致性，并利用这些评估来指导早停决策和推理依据选择。该框架采用基于准则的停止策略和加权多数投票，从而能够在何时停止采样以及选择哪个推理依据方面做出更明智的决策。我们在多个问答数据集上的全面实验表明，RASC 优于现有方法，在保持准确率的同时将样本使用量减少了约 70%。此外，RASC 有助于选择高保真度的推理依据，从而提升大语言模型输出的忠实度。我们的方法有效解决了大语言模型推理任务中效率与准确率之间的权衡问题，为在资源受限环境中更细致、更忠实且更有效地利用大语言模型提供了新的视角。
英文摘要：
Self-Consistency mitigates hallucinations in Large Language Models (LLMs) by sampling multiple reasoning paths,but it lacks a systematic approach to determine the optimal number of samples or select the most faithful rationale. To address this limitation, we introduce Reasoning-Aware Self-Consistency (RASC), a novel framework that enhances sampling efficiency and reasoning faithfulness by dynamically evaluating both outputs and rationales. RASC assesses the quality of reasoning and the consistency of answers for each generated sample, using these assessments to guide early stopping decisions and rationale selection. The framework employs criteria-based stopping and weighted majority voting, enabling more informed choices on when to halt sampling and which rationale to select. Our comprehensive experiments across diverse question-answering datasets demonstrate that RASC outperforms existing methods, reducing sample usage by approximately 70% while maintaining accuracy. Moreover, RASC facilitates the selection of high-fidelity rationales, thereby improving the faithfulness of LLM outputs. Our approach effectively addresses the efficiency-accuracy trade-off in LLM reasoning tasks, offering a new perspective for more nuanced, faithful, and effective utilization of LLMs in resource-constrained environments.
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。
