PRewrite: Prompt Rewriting with Reinforcement Learning
This addresses the time-consuming and sub-optimal nature of prompt engineering for developers and users of LLM applications, though it is incremental as it builds on existing automated prompt optimization techniques.
The paper tackles the problem of inefficient manual prompt engineering for LLM-based applications by proposing PRewrite, an automated method that uses reinforcement learning to rewrite prompts, resulting in improved performance on diverse benchmark datasets.
Prompt engineering is critical for the development of LLM-based applications. However, it is usually done manually in a "trial and error" fashion that can be time consuming, ineffective, and sub-optimal. Even for the prompts which seemingly work well, there is always a lingering question: can the prompts be made better with further modifications? To address these problems, we investigate automated prompt engineering in this paper. Specifically, we propose PRewrite, an automated method to rewrite an under-optimized prompt to a more effective prompt. We instantiate the prompt rewriter using a LLM. The rewriter LLM is trained using reinforcement learning to optimize the performance on a given downstream task. We conduct experiments on diverse benchmark datasets, which demonstrates the effectiveness of PRewrite.