AI安全

论文复现： REFINE: Inversion-Free Backdoor Defense via Model Reprogramming (ICLR 2025) Link: https://arxiv.org/abs/2502.18508 Github Repo: https://github.com/WhitolfChen/REFINE 实验环境 Platform: Openbayes GPU: NVIDIA GeForce RTX 4090 Mem: 80 GB anaconda3 攻击方法(BadNets) 数据集 训练数据使用 CIFAR-10 数据集，共有 60000 个样本，每个样本都是尺寸为 32$\times$32​ 的 RGB 图像，60000个样本被分为 50000 个训练样本和 10000 个测试样本。CIFAR-10 包含 10 个互斥类别，每个类别有 6,000 张图像：飞机（airplane）、汽车（automobile）、鸟（bird）、猫（cat）、鹿（deer）、狗（dog）、青蛙（frog）、马（horse）、船（ship）、卡车（truck）。 ...

1. Introduction Once upon a time in the mystical kingdom of AI, there lived a cunning wolf known for its deceiving ways. This wasn’t any ordinary wolf though; it was an avatar in the vast digital realm of Machine Learning Forest, often identified by the keen eyes of the vigilant MobileNetV2. However, the wolf was not content with its identity and yearned for a transformation - an evasion. It dreamt of an enchanted costume, one that could confuse even the sharpest observers. A costume so convincing that it would allow the wolf to be perceived as something entirely different. ...