更好的零样本推理：角色扮演提示
Better Zero-Shot Reasoning with Role-Play Prompting
AI总结：
本研究提出角色扮演提示方法，在零样本设置下显著提升大型语言模型的推理能力，在AQuA和Last Letter等基准上准确率大幅提升，并优于零样本思维链技术。
摘要：
现代大型语言模型（LLMs）展现出非凡的角色扮演能力，使其不仅能够扮演人类角色，还能模拟非人类实体。这种多功能性使它们能够在各种情境中模拟复杂的人类互动和行为，以及模仿特定的对象或系统。尽管这些能力增强了用户参与度并引入了新颖的交互模式，但角色扮演对LLMs推理能力的影响仍未得到充分探索。在本研究中，我们引入了一种策略性设计的角色扮演提示方法，并在零样本设置下，跨十二个多样化的推理基准对其性能进行了评估。我们的实证结果表明，在大多数数据集上，角色扮演提示始终优于标准的零样本方法。值得注意的是，在使用ChatGPT进行的实验中，AQuA上的准确率从53.5%提升至63.8%，而在Last Letter上则从23.8%提升至84.2%。进一步与零样本思维链（Zero-Shot-CoT）技术（该技术提示模型“逐步思考”）进行比较后，我们的研究表明，角色扮演提示是思维链（CoT）过程更有效的触发器。这凸显了其增强LLMs推理能力的潜力。我们在https://github.com/NKU-HLT/Role-Play-Prompting发布了代码。
英文摘要：
Modern large language models (LLMs) exhibit a remarkable capacity for role-playing, enabling them to embody not only human characters but also non-human entities. This versatility allows them to simulate complex human-like interactions and behaviors within various contexts, as well as to emulate specific objects or systems. While these capabilities have enhanced user engagement and introduced novel modes of interaction, the influence of role-playing on LLMs' reasoning abilities remains underexplored. In this study, we introduce a strategically designed role-play prompting methodology and assess its performance under the zero-shot setting across twelve diverse reasoning benchmarks. Our empirical results illustrate that role-play prompting consistently surpasses the standard zero-shot approach across most datasets. Notably, in experiments conducted using ChatGPT, accuracy on AQuA rises from 53.5% to 63.8%, and on Last Letter from 23.8% to 84.2%.Upon further comparison with the Zero-Shot-CoT technique, which prompts the model to "think step by step", our study demonstrates that role-play prompting acts as a more effective trigger for the CoT process. This highlights its potential to augment the reasoning capabilities of LLMs. We release our code at https://github.com/NKU-HLT/Role-Play-Prompting.
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。
