Structured Confidence Perturbation via Fisher Geometry for Adversarial Training

研究论文 Hugging Face Papers 原文语言:中文

论文摘要(作者报告,未经本站复测):Adversarial training is often plagued by "pseudo-attacks", perturbations that significantly increase the loss value but fail to truly alter the model's confidence structure.

原文摘录

论文摘要(作者报告,未经本站复测):Adversarial training is often plagued by "pseudo-attacks", perturbations that significantly increase the loss value but fail to truly alter the model's confidence structure.

以下是来源提供的简短摘录,完整内容请查看原文。

参考来源

← 返回每日简报