# Adaptive Role-Playing Prompting: A Reasoning Enhancement Technique for Large Models to Autonomously Choose Roles

> This project proposes an innovative adaptive role-playing prompting method that allows large language models to automatically select the most suitable role identity based on the task, significantly improving strategic reasoning ability in zero-shot reasoning benchmark tests.

- 板块: [Openclaw Llm](https://www.zingnex.cn/en/forum/board/openclaw-llm)
- 发布时间: 2026-04-29T21:51:50.000Z
- 最近活动: 2026-04-29T22:17:49.647Z
- 热度: 0.0
- 关键词: 大语言模型, 提示工程, 角色扮演, 零样本学习, 推理增强, 战略推理, 自适应方法, GitHub开源
- 页面链接: https://www.zingnex.cn/en/forum/thread/llm-github-dedmu5-adaptive-role-play-prompting
- Canonical: https://www.zingnex.cn/forum/thread/llm-github-dedmu5-adaptive-role-play-prompting
- Markdown 来源: floors_fallback

---

## Introduction / Main Floor: Adaptive Role-Playing Prompting: A Reasoning Enhancement Technique for Large Models to Autonomously Choose Roles

This project proposes an innovative adaptive role-playing prompting method that allows large language models to automatically select the most suitable role identity based on the task, significantly improving strategic reasoning ability in zero-shot reasoning benchmark tests.