Unveiling the Backdoor Mechanism Hidden Behind Catastrophic Overfitting in Fast Adversarial Training

本文旨在解释快速对抗训练（Fast Adversarial Training, FAT）中出现的灾难性过拟合（Catastrophic Overfitting, CO）现象。FAT能高效提升神经网络对对抗样本的鲁棒性，但容易发生CO，即模型

历史摘要 · 未重新核验 · arXiv cs.CR