演绎束搜索：解码链式思维推理的可演绎依据
Deductive Beam Search: Decoding Deducible Rationale for Chain-of-Thought Reasoning
AI总结：
本文提出演绎束搜索（DBS），将链式思维与演绎推理及逐步束搜索结合，通过验证器检查推理步骤的可演绎性以减少错误累积，并在多种规模和推理任务上显著提升LLMs性能。
摘要：
近期进展通过各种方法显著增强了大语言模型（LLMs）的推理能力，尤其是链式思维（CoT）推理。然而，以往方法未能解决中间步骤中的推理错误，导致错误累积。在本文中，我们提出演绎束搜索（DBS），它将CoT和演绎推理与逐步束搜索无缝集成用于LLMs。我们的方法部署一个验证器，验证推理步骤及其前提的可演绎性，从而缓解错误累积。此外，我们引入一种可扩展且无需人工的数据构建方法，以增强我们模型的验证能力。大量实验表明，我们的方法显著提升了不同规模LLMs（7B、13B、70B和ChatGPT）在8个推理数据集上的基础性能，这些数据集来自3种不同的推理类型，包括算术、常识和符号推理。此外，我们的分析证明了DBS能够检测多样且细微的推理错误，并在不同模型规模上具有鲁棒性。
英文摘要：
Recent advancements have significantly augmented the reasoning capabilities of Large Language Models (LLMs) through various methodologies, especially chain-of-thought (CoT) reasoning. However, previous methods fail to address reasoning errors in intermediate steps, leading to accumulative errors. In this paper, we propose Deductive Beam Search (DBS), which seamlessly integrates CoT and deductive reasoning with step-wise beam search for LLMs. Our approach deploys a verifier, verifying the deducibility of a reasoning step and its premises, thus alleviating the error accumulation. Furthermore, we introduce a scalable and labor-free data construction method to amplify our model's verification capabilities. Extensive experiments demonstrate that our approach significantly enhances the base performance of LLMs of various scales (7B, 13B, 70B, and ChatGPT) across 8 reasoning datasets from 3 diverse reasoning genres, including arithmetic, commonsense, and symbolic. Moreover, our analysis proves DBS's capability of detecting diverse and subtle reasoning errors and robustness on different model scales.
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。
