摘要 攻击者可以通过侧信道信息(Side-channel)完成模型窃取攻击[17]. [17] Hua W Z, Zhang Z R, Suh G E. Reverse Engineering Convolutional Neural Networks through Side-channel Information Leaks[C]. 2018 55th ACM/ESDA/IEEE Des
本文介绍与「Prompt窃取与保护」相关的几篇工作。 目录 1. Prompt Stealing Attacks Against Text-to-Image Generation Models(PromptStealer)2. Hard Prompts Made Easy: Gradient-Based Discrete Optimization for Prompt Tuning