ADOPT: Adaptive Dependency-Guided Joint Prompt Optimization for Multi-Step LLM Pipelines

TL;DR

ADOPT introduces an adaptive, dependency-guided framework for joint prompt optimization in multi-step LLM pipelines, improving performance by analyzing step dependencies and allocating resources effectively.

Contribution

It presents a novel method that analyzes step dependencies, decomposes global gradients, and adaptively allocates optimization resources, enhancing multi-step prompt tuning.

Findings

ADOPT outperforms baseline prompt optimization methods on real-world datasets.

It demonstrates robustness across diverse pipeline structures.

The framework effectively leverages step dependencies for improved prompt updates.

Abstract

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.