arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2304.03472cs.CR
收藏

提示调优语言模型能否确保隐私?

Does Prompt-Tuning Language Model Ensure Privacy?

Shangyu Xie, Wei Dai, Esha Ghosh, Sambuddha Roy, Dan Schwartz, Kim Laine

AI总结:

本文首次研究提示调优场景下的隐私泄露问题,提出利用用户特定信号推断私有信息的隐私攻击框架,并通过真实邮件服务流程验证了提示调优的泄露风险及攻击有效性。

摘要:

提示调优作为语言领域中一种高效的调优方法受到关注,即调优一个仅由少量token组成的提示,同时保持大型语言模型冻结,却能取得与传统微调相当的性能。考虑到语言模型日益突出的隐私问题,我们首次在提示调优场景下开展隐私泄露研究。我们首先描述了一个真实的电子邮件服务流程,该流程通过提示调优为不同用户提供定制化输出。随后,我们提出一种新颖的隐私攻击框架,利用带有用户特定信号的提示模块来推断用户的私有信息。我们在目标流程上进行了全面的隐私评估,以展示提示调优可能带来的泄露风险。实验结果同时证明了所提攻击方法的有效性。

英文摘要:

Prompt-tuning has received attention as an efficient tuning method in the language domain, i.e., tuning a prompt that is a few tokens long, while keeping the large language model frozen, yet achieving comparable performance with conventional fine-tuning. Considering the emerging privacy concerns with language models, we initiate the study of privacy leakage in the setting of prompt-tuning. We first describe a real-world email service pipeline to provide customized output for various users via prompt-tuning. Then we propose a novel privacy attack framework to infer users' private information by exploiting the prompt module with user-specific signals. We conduct a comprehensive privacy evaluation on the target pipeline to demonstrate the potential leakage from prompt-tuning. The results also demonstrate the effectiveness of the proposed attack.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑