Recent advancements in large language models have showcased their remarkable generalizability across various domains. However, their reasoning abilities still have significant room for improvement, especially when confronted with scenarios requiring multi-step reasoning. Although large language models possess extensive knowledge, their reasoning often fails to effectively utilize this knowledge to establish a coherent thinking paradigm. These models sometimes show hallucinations as their reasoning procedures are unconstrained by logical principles. Aiming at improving the zero-shot chain-of-thought reasoning ability of large language models, we propose LoT (Logical Thoughts) prompting, a self-improvement framework that leverages principles rooted in symbolic logic, particularly Reductio ad Absurdum, to systematically verify and rectify the reasoning processes step by step. Experimental evaluations conducted on language tasks in diverse domains, including arithmetic, commonsense, symbolic, causal inference, and social problems, demonstrate the efficacy of enhanced reasoning by logic.

翻译：近期大型语言模型的进展展示了其在各个领域的显著泛化能力。然而，其推理能力仍有较大提升空间，尤其是在面对需要多步推理的场景时。尽管大型语言模型具备广泛的知识，但其推理过程往往未能有效利用这些知识来构建连贯的思维范式。由于推理过程缺乏逻辑原则的约束，这些模型有时会出现幻觉现象。为提升大型语言模型的零样本思维链推理能力，我们提出逻辑思维（Logical Thoughts, LoT）提示方法——一种自我改进框架，该框架利用符号逻辑（特别是归谬法）的原则，逐步系统地验证和修正推理过程。在算术、常识、符号推理、因果推理及社会问题等多个领域的语言任务上开展的实验评估表明，通过逻辑增强的推理具有有效性。