基于标注数据的思维链自动提示增强与选择
Automatic Prompt Augmentation and Selection with Chain-of-Thought from Labeled Data
AI总结：
本文提出Automate-CoT策略，通过从少量标注数据自动增广并剪枝推理链、再用方差缩减策略梯度选择最优组合，实现无需人工标注推理链的思维链提示，并在多类推理任务上取得提升。
摘要：
思维链（CoT）提升了大语言模型（LLMs）的推理能力，并在复杂推理任务中取得了优越性能。然而，大多数CoT研究依赖精心设计的人工标注推理链来提示LLMs，这给现实应用中仅有标注数据而缺乏推理链的场景带来了挑战。本文提出一种新策略 Automate-CoT（基于思维链的自动提示增强与选择），该策略能够绕过人工构建CoT的过程：首先从一个小规模标注数据集中自动增广推理链，然后基于标签剪除低质量推理链，构建机器生成的推理链候选池。最后，通过采用方差缩减的策略梯度策略估计每个样本的重要性，从候选池中选出若干推理链的最优组合用于CoT提示。Automate-CoT使CoT技术能够快速适配不同任务。实验结果表明了该方法的有效性，在算术推理（+2.7%）、常识推理（+3.4%）、符号推理（+3.2%）和非推理任务（+2.5%）上均取得了具有竞争力的结果。代码可在 https://github.com/SHUMKASHUN/Automate-CoT 获取。
英文摘要：
Chain-of-thought (CoT) advances the reasoning abilities of large language models (LLMs) and achieves superior performance in complex reasoning tasks. However, most CoT studies rely on carefully designed human-annotated rational chains to prompt LLMs, posing challenges for real-world applications where labeled data is available without rational chains. This paper proposes a new strategy, Automate-CoT (Automatic Prompt Augmentation and Selection with Chain-of-Thought), that can bypass human engineering of CoT by automatically augmenting rational chains from a small labeled dataset, and then pruning low-quality chains to construct a candidate pool of machine-generated rationale chains based on the labels. Finally, it selects the optimal combination of several rationale chains from the pool for CoT prompting by employing a variance-reduced policy gradient strategy to estimate the significance of each example. Automate-CoT enables a quick adaptation of the CoT technique to different tasks. Experimental results demonstrate the effectiveness of our method, where competitive results are achieved on arithmetic reasoning (+2.7%), commonsense reasoning (+3.4%), symbolic reasoning (+3.2%), and non-reasoning tasks (+2.5%). The code is available at https://github.com/SHUMKASHUN/Automate-CoT.
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。
