Gradient-based Adversarial Attacks against Text Transformers

Code

• github.com/facebookresearch/text-adversarial-attack
  OfficialIn paperpytorch★ 110
• github.com/gmum/mango
  pytorch★ 4

Abstract

We propose the first general-purpose gradient-based attack against transformer models. Instead of searching for a single adversarial example, we search for a distribution of adversarial examples parameterized by a continuous-valued matrix, hence enabling gradient-based optimization. We empirically demonstrate that our white-box attack attains state-of-the-art attack performance on a variety of natural language tasks. Furthermore, we show that a powerful black-box transfer attack, enabled by sampling from the adversarial distribution, matches or exceeds existing methods, while only requiring hard-label outputs.

Reproductions