Improving the reasoning capabilities of large language models (LLMs) has attracted considerable interest. Recent approaches primarily focus on improving the reasoning process to yield a more precise final answer. However, in scenarios involving contextually aware reasoning, these methods neglect the importance of first identifying logical relationships from the context before proceeding with the reasoning. This oversight could lead to a superficial understanding and interaction with the context, potentially undermining the quality and reliability of the reasoning outcomes. In this paper, we propose an information re-organization (InfoRE) method before proceeding with the reasoning to enhance the reasoning ability of LLMs. We first perform a re-organization processing of the contextual content, e.g., documents or paragraphs, to obtain logical relationships. Then, we utilize the re-organized information in the reasoning process. This enables LLMs to deeply understand the contextual content by clearly perceiving these logical relationships. To demonstrate the effectiveness of our approach in improving the reasoning ability, we conduct experiments using Llama2-70B, GPT-3.5, and GPT-4 on various contextually aware multi-hop reasoning tasks. Using only a zero-shot setting, our method achieves an average improvement of 3\% across all tasks, highlighting its potential to improve the reasoning performance of LLMs. Our source code is available at https://github.com/hustcxx/InfoRE.

翻译：提升大型语言模型（LLMs）的推理能力已引起广泛关注。近期方法主要侧重于优化推理过程，以得出更精确的最终答案。然而，在涉及上下文感知推理的场景中，这些方法忽略了在推理前首先从上下文中识别逻辑关系的重要性。这种疏忽可能导致对上下文的浅层理解与交互，从而可能损害推理结果的质量与可靠性。本文提出一种在推理前进行信息重组（InfoRE）的方法，以增强LLMs的推理能力。我们首先对文档或段落等上下文内容进行重组处理，获取其中的逻辑关系；随后，将重组后的信息用于推理过程。这使LLMs能够通过清晰感知这些逻辑关系，深入理解上下文内容。为证明该方法在提升推理能力方面的有效性，我们基于Llama2-70B、GPT-3.5和GPT-4在多种上下文感知的多跳推理任务上开展实验。在仅使用零样本设置的情况下，该方法在所有任务上平均提升3%，凸显了其改善LLMs推理性能的潜力。我们的源代码已开源至https://github.com/hustcxx/InfoRE。