标题： PR$εε$MPT：大型语言模型的敏感提示清洗 显示英文标题

标题： Pr$εε$mpt: Sanitizing Sensitive Prompts for LLMs

摘要： 大型语言模型（LLMs）的兴起带来了新的隐私挑战，特别是在推理阶段，提示中的敏感信息可能会暴露给专有LLM API。 本文解决了在保持响应质量的同时正式保护提示中包含的敏感信息的问题。 为此，首先，我们引入了一种受密码学启发的提示 sanitization 概念，该概念将输入提示转换以保护其敏感标记。 其次，我们提出了 Pr$\epsilon\epsilon$mpt，一种实现提示 sanitization 的新颖系统。 Pr$\epsilon\epsilon$mpt 将敏感标记分类为两类：(1) 那些LLM的响应仅取决于格式的标记（如SSN、信用卡号码），对于这些标记我们使用保留格式的加密（FPE）；(2)那些响应依赖于特定值的标记（如年龄、工资），对于这些标记我们应用度量差分隐私（mDP）。 我们的评估表明，Pr$\epsilon\epsilon$mpt 是一种实用的方法，在保持高实用性的同时实现有意义的隐私保证，并且优于先前的方法。 显示英文摘要

摘要： The rise of large language models (LLMs) has introduced new privacy challenges, particularly during inference where sensitive information in prompts may be exposed to proprietary LLM APIs. In this paper, we address the problem of formally protecting the sensitive information contained in a prompt while maintaining response quality. To this end, first, we introduce a cryptographically inspired notion of a prompt sanitizer which transforms an input prompt to protect its sensitive tokens. Second, we propose Pr$\epsilon\epsilon$mpt, a novel system that implements a prompt sanitizer. Pr$\epsilon\epsilon$mpt categorizes sensitive tokens into two types: (1) those where the LLM's response depends solely on the format (such as SSNs, credit card numbers), for which we use format-preserving encryption (FPE); and (2) those where the response depends on specific values, (such as age, salary) for which we apply metric differential privacy (mDP). Our evaluation demonstrates that Pr$\epsilon\epsilon$mpt is a practical method to achieve meaningful privacy guarantees, while maintaining high utility compared to unsanitized prompts, and outperforming prior methods