Deductive Beam Search: Decoding Deducible Rationale for Chain-of-Thought Reasoning

Tinghui Zhu, Kai Zhang, Jian Xie, Yu Su

Research Area: Inference algorithms for LMs

Keywords: Deductive Reasoning, Large Language Models, Decoding Algorithm

TL;DR: Deductive Beam Search, constrained by the principle of deductive reasoning, integrates CoT with step-wise beam search to guide reasoning towards a deducible path.

Abstract: 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 often struggle to address reasoning errors in intermediate steps, which can lead 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.

Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the COLM Code of Ethics on https://colmweb.org/CoE.html

Author Guide: I certify that this submission complies with the submission instructions as described on https://colmweb.org/AuthorGuide.html

Submission Number: 399