SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2002.11242 · ICML · 2020

Attacks Which Do Not Kill Training Make Adversarial Learning Stronger

Masashi Sugiyama, Gang Niu, Jingfeng Zhang, Xilie Xu, Bo Han, Lizhen Cui, Mohan Kankanhalli

arXiv · PDF · Open in the Atlas

Code that ran

We have not lifted any functions out of this paper's repositories yet, so there is nothing we have run. The repositories linked to it are listed below.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Adversarial training based on the minimax formulation is necessary for obtaining adversarial robustness of trained models. However, it is conservative or even pessimistic so that it sometimes hurts the natural generalization. In this paper, we raise a fundamental question-do we have to trade off natural generalization for adversarial robustness? We argue that adversarial training is to employ confident adversarial data for updating the current model. We propose a novel formulation of friendly adversarial training (FAT): rather than employing most adversarial data maximizing the loss, we search for least adversarial data (i.e., friendly adversarial data) minimizing the loss, among the adversarial data that are confidently misclassified. Our novel formulation is easy to implement by just stopping the most adversarial data searching algorithms such as PGD (projected gradient descent) early, which we call early-stopped PGD. Theoretically, FAT is justified by an upper bound of the adversarial risk. Empirically, early-stopped PGD allows us to answer the earlier question negatively-adversarial robustness can indeed be achieved without compromising the natural generalization.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2002.11242")
get_code_for_paper("2002.11242")
have("2002.11242")

The run record, dated, one paper per request, free:

curl https://syntology.ai/api/ran/2002.11242.json

A badge for a README (the split and the date, never a ratio):

[![Syntology run record](https://syntology.ai/api/ran/2002.11242.svg)](https://syntology.ai/paper/2002.11242)

Connect an agent — have() is free.