Helix: A Dual-Helix Co-Evolutionary Multi-Agent System for Prompt Optimization and Question Reformulation

TL;DR

Helix introduces a co-evolutionary multi-agent system that jointly optimizes question reformulation and prompt instructions, significantly enhancing large language model performance across multiple benchmarks.

Contribution

The paper presents a novel unified framework that simultaneously refines questions and prompts through a structured three-stage co-evolution process, addressing limitations of existing methods.

Findings

Achieves up to 3.95% performance improvement on 12 benchmarks.

Demonstrates effectiveness over 6 strong baseline methods.

Shows improved optimization efficiency in prompt and question refinement.

Abstract

Automated prompt optimization (APO) aims to improve large language model performance by refining prompt instructions. However, existing methods are largely constrained by fixed prompt templates, limited search spaces, or single-sided optimization that treats user questions as immutable inputs. In practice, question formulation and prompt design are inherently interdependent: clearer question structures facilitate focused reasoning and task understanding, while effective prompts reveal better ways to organize and restate queries. Ignoring this coupling fundamentally limits the effectiveness and adaptability of current APO approaches. We propose a unified multi-agent system (Helix) that jointly optimizes question reformulation and prompt instructions through a structured three-stage co-evolutionary framework. Helix integrates (1) planner-guided decomposition that breaks optimization into coupled question-prompt objectives, (2) dual-track co-evolution where specialized agents iteratively refine and critique each other to produce complementary improvements, and (3) strategy-driven question generation that instantiates high-quality reformulations for robust inference. Extensive experiments on 12 benchmarks against 6 strong baselines demonstrate the effectiveness of Helix, achieving up to 3.95% performance improvements across tasks with favorable optimization efficiency.