语言模型在含噪声推理路径的思维链提示下能否实现鲁棒推理？
Can Language Models Perform Robust Reasoning in Chain-of-thought Prompting with Noisy Rationales?
AI总结：
针对大语言模型思维链提示中噪声推理路径导致准确率下降的问题，提出仅需一条干净推理路径的对比去噪方法CD-CoT，在NoRa数据集上较基础模型平均准确率提升17.8%，去噪能力优于现有基线方法。
摘要：
本文研究了大语言模型（LLMs）中一个尚未被充分探索的挑战：含噪声推理路径的思维链提示，即上下文学习所用示例中包含无关或不准确的推理思路。我们构建了专门用于评估存在噪声推理路径时推理鲁棒性的NoRa数据集。在NoRa数据集上的研究结果表明，当前大语言模型普遍易受此类噪声影响，而自校正、自一致性等现有鲁棒性方法的效果十分有限。值得注意的是，与使用干净推理路径的提示相比，基础大语言模型在存在无关思路时准确率下降1.4%-19.8%，在存在不准确思路时降幅更大，达2.2%-40.4%。\n 应对这一挑战需要实际可获取的外部监督。为此，我们提出了带噪声思维链的对比去噪方法（CD-CoT）。该方法仅需一条干净推理路径（这可作为去噪类提示的最低要求），通过将其与噪声推理路径进行对比，增强大语言模型的去噪推理能力。该方法遵循探索与利用原则：（1）在输入空间中对推理路径进行改写和筛选，实现显式去噪；（2）在输出空间中探索多样推理路径并对答案进行投票。实验表明，CD-CoT的准确率较基础模型平均提升17.8%，且去噪能力显著强于基线方法。源代码已公开于：https://github.com/tmlr-group/NoisyRationales。
英文摘要：
This paper investigates an under-explored challenge in large language models (LLMs): chain-of-thought prompting with noisy rationales, which include irrelevant or inaccurate reasoning thoughts within examples used for in-context learning. We construct NoRa dataset that is tailored to evaluate the robustness of reasoning in the presence of noisy rationales. Our findings on NoRa dataset reveal a prevalent vulnerability to such noise among current LLMs, with existing robust methods like self-correction and self-consistency showing limited efficacy. Notably, compared to prompting with clean rationales, base LLM drops by 1.4%-19.8% in accuracy with irrelevant thoughts and more drastically by 2.2%-40.4% with inaccurate thoughts. Addressing this challenge necessitates external supervision that should be accessible in practice. Here, we propose the method of contrastive denoising with noisy chain-of-thought (CD-CoT). It enhances LLMs' denoising-reasoning capabilities by contrasting noisy rationales with only one clean rationale, which can be the minimal requirement for denoising-purpose prompting. This method follows a principle of exploration and exploitation: (1) rephrasing and selecting rationales in the input space to achieve explicit denoising and (2) exploring diverse reasoning paths and voting on answers in the output space. Empirically, CD-CoT demonstrates an average improvement of 17.8% in accuracy over the base model and shows significantly stronger denoising capabilities than baseline methods. The source code is publicly available at: https://github.com/tmlr-group/NoisyRationales.
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。
