让大模型多步任务更准：通过依赖分析精准优化每一步提示词
ADOPT: Adaptive Dependency-Guided Joint Prompt Optimization for Multi-Step LLM Pipelines
- 基于步骤间依赖关系，从最终错误反推每步的优化信号
- 在真实数据集上比现有方法提升12%以上准确率，效果稳定
- 适合需要多步推理的大模型应用，如复杂问答与规划
多步大模型流水线可解决复杂任务，但因缺乏步骤级监督和步骤间依赖关系，联合优化提示词仍具挑战。本文提出 ADOPT，一种自适应依赖引导的联合提示优化框架。该方法分析每步模型输出与最终结果间的依赖关系，从最终任务误差构建全局文本梯度，并分解为各步骤的局部文本梯度，提供更精确的本地提示更新信号。同时，解耦信号估计与提示更新，支持灵活集成单提示优化器，并采用基于 Shapley 值的策略自适应分配优化资源至高影响步骤。在真实世界数据集和结构多样的流水线上实验表明，ADOPT 效果显著且鲁棒，持续优于多个强基线方法。
原文摘要 · Abstract (English)
Multi-step LLM pipelines can solve complex tasks, but jointly optimizing prompts across steps remains challenging due to missing step-level supervision and inter-step dependency. We propose ADOPT, an adaptive dependency-guided joint prompt optimization framework for multi-step LLM pipelines. ADOPT analyzes the dependency between each LLM step and the final output, constructs a global textual gradient from final-task errors, and decomposes it into step-level local textual gradients, providing more precise optimization signals for local prompt updates. It further decouples signal estimation from prompt updating, enabling flexible integration of single-prompt optimizers, and uses a Shapley-based strategy to adaptively allocate optimization resources to high-impact steps. Experiments on real-world datasets and structurally diverse pipelines demonstrate that ADOPT is effective and robust, consistently outperforming strong prompt optimization baselines.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。