SI-Agent: An Agentic Framework for Feedback-Driven Generation and Tuning of Human-Readable System Instructions for Large Language Models

TL;DR

SI-Agent is a novel framework that automatically generates and refines human-readable system instructions for LLMs using an agentic, feedback-driven process, improving interpretability and performance.

Contribution

The paper introduces SI-Agent, an innovative agentic framework that automates the creation and iterative refinement of human-readable system instructions for LLMs, combining multiple agents and feedback loops.

Findings

SI-Agent produces effective, readable system instructions.

It achieves a better balance between task performance and interpretability.

The framework demonstrates efficiency in instruction tuning.

Abstract

System Instructions (SIs), or system prompts, are pivotal for guiding Large Language Models (LLMs) but manual crafting is resource-intensive and often suboptimal. Existing automated methods frequently generate non-human-readable "soft prompts," sacrificing interpretability. This paper introduces SI-Agent, a novel agentic framework designed to automatically generate and iteratively refine human-readable SIs through a feedback-driven loop. SI-Agent employs three collaborating agents: an Instructor Agent, an Instruction Follower Agent (target LLM), and a Feedback/Reward Agent evaluating task performance and optionally SI readability. The framework utilizes iterative cycles where feedback guides the Instructor's refinement strategy (e.g., LLM-based editing, evolutionary algorithms). We detail the framework's architecture, agent roles, the iterative refinement process, and contrast it with existing methods. We present experimental results validating SI-Agent's effectiveness, focusing on metrics for task performance, SI readability, and efficiency. Our findings indicate that SI-Agent generates effective, readable SIs, offering a favorable trade-off between performance and interpretability compared to baselines. Potential implications include democratizing LLM customization and enhancing model transparency. Challenges related to computational cost and feedback reliability are acknowledged.