DORY: Deliberative Prompt Recovery for LLM

TL;DR

DORY is a novel method that leverages output uncertainty to accurately recover prompts from limited outputs of LLMs, improving performance and establishing new benchmarks without external resources.

Contribution

We introduce DORY, a cost-effective prompt recovery approach that uses uncertainty to reconstruct prompts from limited outputs, outperforming existing methods.

Findings

DORY improves prompt recovery performance by approximately 10.82%.

DORY establishes a new state-of-the-art in prompt recovery tasks.

DORY operates effectively using only a single LLM without external resources.

Abstract

Prompt recovery in large language models (LLMs) is crucial for understanding how LLMs work and addressing concerns regarding privacy, copyright, etc. The trend towards inference-only APIs complicates this task by restricting access to essential outputs for recovery. To tackle this challenge, we extract prompt-related information from limited outputs and identify a strong(negative) correlation between output probability-based uncertainty and the success of prompt recovery. This finding led to the development of Deliberative PrOmpt RecoverY (DORY), our novel approach that leverages uncertainty to recover prompts accurately. DORY involves reconstructing drafts from outputs, refining these with hints, and filtering out noise based on uncertainty. Our evaluation across diverse LLMs and prompt benchmarks shows that DORY outperforms existing baselines, improving performance by approximately 10.82% and establishing a new state-of-the-art record in prompt recovery tasks. Significantly, DORY operates using a single LLM without any external resources or model, offering a cost-effective, user-friendly prompt recovery solution.