Self-Prompt Tuning: Enable Autonomous Role-Playing in LLMs

TL;DR

This paper introduces self-prompt tuning, a method where LLMs generate their own role prompts through fine-tuning, leading to improved performance on NLP benchmarks without manual prompt design.

Contribution

The authors propose a novel self-prompt tuning approach that enables LLMs to autonomously generate role prompts via fine-tuning, reducing manual effort and enhancing performance.

Findings

Self-prompt tuned LLMs outperform instruction-tuned baselines on multiple NLP benchmarks.

The method automates complex prompt design, making LLMs more autonomous.

The approach is validated on Llama-2-7B and Mistral-7B models.

Abstract

Recent advancements in LLMs have showcased their remarkable role-playing capabilities, able to accurately simulate the dialogue styles and cognitive processes of various roles based on different instructions and contexts. Studies indicate that assigning LLMs the roles of experts, a strategy known as role-play prompting, can enhance their performance in the corresponding domains. However, the prompt needs to be manually designed for the given problem, requiring certain expertise and iterative modifications. To this end, we propose self-prompt tuning, making LLMs themselves generate role-play prompts through fine-tuning. Leveraging the LIMA dataset as our foundational corpus, we employ GPT-4 to annotate role-play prompts for each data points, resulting in the creation of the LIMA-Role dataset. We then fine-tune LLMs like Llama-2-7B and Mistral-7B on LIMA-Role. Consequently, the self-prompt tuned LLMs can automatically generate expert role prompts for any given question. We extensively evaluate self-prompt tuned LLMs on widely used NLP benchmarks and open-ended question test. Our empirical results illustrate that self-prompt tuned LLMs outperform standard instruction tuned baselines across most datasets. This highlights the great potential of utilizing fine-tuning to enable LLMs to self-prompt, thereby automating complex prompting strategies. We release the dataset, models, and code at this \href{https://anonymous.4open.science/r/Self-Prompt-Tuning-739E/}{url}.