让大模型自我纠错，一步步更准地回答复杂问题。
Self-Critique Guided Iterative Reasoning for Multi-hop Question Answering
- 通过自省反馈引导迭代推理，动态优化每步思考。
- 在三个数据集上超越现有最佳方法8.6%准确率。
- 适合需要深度逻辑推理的问答系统研究者使用。
尽管大语言模型在推理方面表现出色，但在知识密集型多跳问答任务中仍存在挑战。现有方法虽采用迭代检索，但缺乏中间过程指导，常导致检索错误与推理偏差。为此，我们提出自省引导的迭代推理（SiGIR），通过端到端训练使模型能够通过问题分解逐步求解，并自评估中间推理步骤。在迭代过程中，模型进行分支探索，并利用自评结果选择更有前景的推理路径。在三个多跳推理数据集上的实验表明，该方法显著优于当前最优模型，提升达8.6%。深入分析也为未来研究提供了洞见。代码、数据和模型已开源于GitHub：https://github.com/zchuz/SiGIR-MHQA。
原文摘要 · Abstract (English)
Although large language models (LLMs) have demonstrated remarkable reasoning capabilities, they still face challenges in knowledge-intensive multi-hop reasoning. Recent work explores iterative retrieval to address complex problems. However, the lack of intermediate guidance often results in inaccurate retrieval and flawed intermediate reasoning, leading to incorrect reasoning. To address these, we propose Self-Critique Guided Iterative Reasoning (SiGIR), which uses self-critique feedback to guide the iterative reasoning process. Specifically, through end-to-end training, we enable the model to iteratively address complex problems via question decomposition. Additionally, the model is able to self-evaluate its intermediate reasoning steps. During iterative reasoning, the model engages in branching exploration and employs self-evaluation to guide the selection of promising reasoning trajectories. Extensive experiments on three multi-hop reasoning datasets demonstrate the effectiveness of our proposed method, surpassing the previous SOTA by $8.6\%$. Furthermore, our thorough analysis offers insights for future research. Our code, data, and models are available at Github: https://github.com/zchuz/SiGIR-MHQA.
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。