DORY: Deliberative Prompt Recovery for LLM

Lirong Gao, Ru Peng, Yiming Zhang, Junbo Zhao


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 shows that DORY outperforms existing baselines across diverse LLMs and prompt benchmarks, 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.
Anthology ID:
2024.findings-acl.631
Volume:
Findings of the Association for Computational Linguistics: ACL 2024
Month:
August
Year:
2024
Address:
Bangkok, Thailand
Editors:
Lun-Wei Ku, Andre Martins, Vivek Srikumar
Venue:
Findings
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
10614–10632
Language:
URL:
https://aclanthology.org/2024.findings-acl.631
DOI:
10.18653/v1/2024.findings-acl.631
Bibkey:
Cite (ACL):
Lirong Gao, Ru Peng, Yiming Zhang, and Junbo Zhao. 2024. DORY: Deliberative Prompt Recovery for LLM. In Findings of the Association for Computational Linguistics: ACL 2024, pages 10614–10632, Bangkok, Thailand. Association for Computational Linguistics.
Cite (Informal):
DORY: Deliberative Prompt Recovery for LLM (Gao et al., Findings 2024)
Copy Citation:
PDF:
https://aclanthology.org/2024.findings-acl.631.pdf