Adaptive Prompt Structure Factorization: A Framework for Self-Discovering and Optimizing Compositional Prompt Programs

TL;DR

Adaptive Prompt Structure Factorization (aPSF) is a framework that automatically discovers and optimizes compositional prompt structures for large language models, improving reasoning accuracy and efficiency.

Contribution

It introduces a novel API-only method that identifies task-specific prompt components and performs targeted, sample-efficient updates based on their contribution.

Findings

aPSF outperforms strong baselines on reasoning benchmarks

it improves accuracy by up to 2.16 percentage points

reduces optimization tokens by 45-87% on MultiArith

Abstract

Automated prompt optimization is crucial for eliciting reliable reasoning from large language models (LLMs), yet most API-only prompt optimizers iteratively edit monolithic prompts, coupling components and obscuring credit assignment, limiting controllability, and wasting tokens. We propose Adaptive Prompt Structure Factorization (aPSF), an API-only framework (prompt-in/text-out; no access to model internals) that uses an Architect model to discover task-specific prompt structures as semantic factors. aPSF then performs interventional, single-factor updates: interventional factor-level scoring estimates each factor's marginal contribution via validation-performance changes, and error-guided factor selection routes updates to the current dominant failure source for more sample-efficient optimization. Across multiple advanced reasoning benchmarks, aPSF outperforms strong baselines including principle-aware optimizers, improving accuracy by up to +2.16 percentage points on average, and reduces optimization cost by 45--87% tokens on MultiArith while reaching peak validation in 1 step.