arXiv 2501.020262025-01-03cs.CLcs.AIcs.LGcs.LO

逻辑思维的递归分解：面向大语言模型卓越推理与知识传播的框架

Recursive Decomposition of Logical Thoughts: Framework for Superior Reasoning and Knowledge Propagation in Large Language Models

Southwest Jiaotong University（西南交通大学）；University of Jeddah（吉达大学）；National Textile University（国立纺织大学）

Kaleem Ullah Qasim, Jiashu Zhang, Tariq Alsahfi, Ateeq Ur Rehman Butt

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

本文提出 RDoLT 提示框架，通过递归任务分解、思维选择评分及强弱思维知识传播提升 LLM 推理，并在多个数学与逻辑基准上显著超越现有方法。

摘要：

增强大语言模型的推理能力仍是人工智能领域的一项关键挑战。我们提出 RDoLT（Recursive Decomposition of Logical Thought prompting，逻辑思维递归分解提示），这是一个可显著提升 LLM 推理性能的新颖框架。RDoLT 建立在三项关键创新之上：（1）将复杂推理任务递归分解为复杂度递进的子任务；（2）采用先进的选择与评分机制，以识别最有前景的推理思维；（3）集成知识传播模块，该模块通过记录强思维和弱思维以进行信息传播，从而模拟人类学习。我们的方法在多个基准上进行了评估，包括 GSM8K、SVAMP、MultiArith、LastLetterConcatenation 和 Gaokao2023 Math。结果表明，RDoLT 始终优于现有最先进技术：在使用 ChatGPT-4 时，其在 GSM8K 上达到 90.98% 的准确率，超过最先进技术 6.28%。在其他基准上也观察到类似提升，准确率增幅为 5.5% 至 6.75%。这些发现凸显了 RDoLT 推动提示工程发展的潜力，为复杂推理任务提供了一种更有效、更具泛化性的方法。

英文摘要：

Enhancing the reasoning capabilities of Large Language Models remains a critical challenge in artificial intelligence. We introduce RDoLT, Recursive Decomposition of Logical Thought prompting, a novel framework that significantly boosts LLM reasoning performance. RDoLT is built on three key innovations: (1) recursively breaking down complex reasoning tasks into sub-tasks of progressive complexity; (2) employing an advanced selection and scoring mechanism to identify the most promising reasoning thoughts; and (3) integrating a knowledge propagation module that mimics human learning by keeping track of strong and weak thoughts for information propagation. Our approach was evaluated across multiple benchmarks, including GSM8K, SVAMP, MultiArith, LastLetterConcatenation, and Gaokao2023 Math. The results demonstrate that RDoLT consistently outperforms existing state-of-the-art techniques, achieving a 90.98 percent accuracy on GSM8K with ChatGPT-4, surpassing state-of-the-art techniques by 6.28 percent. Similar improvements were observed on other benchmarks, with accuracy gains ranging from 5.5 percent to 6.75 percent. These findings highlight RDoLT's potential to advance prompt engineering, offering a more effective and generalizable approach to complex reasoning tasks.

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。