BriefGPT.xyz
Nov, 2023
模型压缩与对抗鲁棒性的关系：当前证据综述
Relationship between Model Compression and Adversarial Robustness: A Review of Current Evidence
HTML
PDF
Svetlana Pavlitska, Hannes Grolig, J. Marius Zöllner
TL;DR
增加模型容量是增强深度学习网络对抗性鲁棒性的已知方法之一，而剪枝和量化等压缩技术则可以减小网络的大小同时保持准确性。本研究总结了现有证据并讨论了观察到的效果可能的解释。
Abstract
Increasing the
model capacity
is a known approach to enhance the
adversarial robustness
of deep learning networks. On the other hand, various
mod
→