通过打乱前提和推理步骤顺序，提升大模型逻辑推理的泛化能力。
Order Doesn't Matter, But Reasoning Does: Training LLMs with Order-Centric Augmentation
- 基于逻辑可交换性，随机打乱独立前提顺序作为数据增强。
- 构建有向无环图建模推理步骤依赖，识别合法重排方式。
- 在多个推理基准上显著提升模型对不同逻辑结构的适应性。
逻辑推理对大语言模型（LLMs）实现准确、连贯的推断至关重要。然而，现有模型在面对推理顺序变化时表现不佳，难以泛化到逻辑等价的变换形式，往往依赖固定的序列模式而非真正的逻辑理解。为此，本文提出一种以顺序为中心的数据增强框架，利用逻辑推理中的可交换性特性。首先，通过随机打乱独立前提引入条件顺序增强；对于推理步骤，构建有向无环图（DAG）以建模步骤间的依赖关系，从而识别在保持逻辑正确性的前提下有效的步骤重排。该方法使模型发展出更灵活、泛化的推理机制。我们在多个逻辑推理基准上进行了广泛实验，结果表明该方法显著提升了模型的推理性能与对多样化逻辑结构的适应能力。代码与增强数据已公开于 https://github.com/qianxiHe147/Order-Centric-Data-Augmentation。
原文摘要 · Abstract (English)
Logical reasoning is essential for large language models (LLMs) to ensure accurate and coherent inference. However, LLMs struggle with reasoning order variations and fail to generalize across logically equivalent transformations. LLMs often rely on fixed sequential patterns rather than true logical understanding. To address this issue, we introduce an order-centric data augmentation framework based on commutativity in logical reasoning. We first randomly shuffle independent premises to introduce condition order augmentation. For reasoning steps, we construct a directed acyclic graph (DAG) to model dependencies between steps, which allows us to identify valid reorderings of steps while preserving logical correctness. By leveraging order-centric augmentations, models can develop a more flexible and generalized reasoning process. Finally, we conduct extensive experiments across multiple logical reasoning benchmarks, demonstrating that our method significantly enhances LLMs' reasoning performance and adaptability to diverse logical structures. We release our codes and augmented data in https://github.com/qianxiHe147/Order-Centric-Data-Augmentation.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。