SCULPT: Systematic Tuning of Long Prompts

TL;DR

SCULPT introduces a hierarchical, tree-based framework for systematically refining long prompts for large language models, improving robustness, interpretability, and performance without requiring initial human prompts.

Contribution

The paper presents SCULPT, a novel hierarchical prompt refinement method that effectively optimizes long prompts, outperforming existing techniques in robustness and interpretability.

Findings

SCULPT improves LLM performance on long prompts.

SCULPT is robust to adversarial perturbations.

SCULPT generates high-quality prompts without initial human input.

Abstract

Prompt optimization is essential for effective utilization of large language models (LLMs) across diverse tasks. While existing optimization methods are effective in optimizing short prompts, they struggle with longer, more complex ones, often risking information loss and being sensitive to small perturbations. To address these challenges, we propose SCULPT (Systematic Tuning of Long Prompts), a framework that treats prompt optimization as a hierarchical tree refinement problem. SCULPT represents prompts as tree structures, enabling targeted modifications while preserving contextual integrity. It employs a Critic-Actor framework that generates reflections and applies actions to refine the prompt. Evaluations demonstrate SCULPT's effectiveness on long prompts, its robustness to adversarial perturbations, and its ability to generate high-performing prompts even without any initial human-written prompt. Compared to existing state of the art methods, SCULPT consistently improves LLM performance by preserving essential task information while applying structured refinements. Both qualitative and quantitative analyses show that SCULPT produces more stable and interpretable prompt modifications, ensuring better generalization across tasks.