Adversarial attacks and backdoor attacks harness task-irrelevant features of datain their implementation . Text style is a feature that is naturally irrelevant to most NLP tasks, and thus suitable for adversarial attacks . The attack success rates can exceed 90% without much effort . It reflects the limitedability of NLP models to handle the feature of text style that has not been realized . The style transfer-based adversarial and backdoor attack methods show superiority to baselines in many aspects . All the code and data of this paper can be obtained at https://://://github.com/thunlp/StyleAttack. All of the data and code and code can be found at http://www.geneanlp/styleAttack.org/styleattack. All the codes and data can be downloaded from Genean

Author(s) : Fanchao Qi, Yangyi Chen, Xurui Zhang, Mukai Li, Zhiyuan Liu, Maosong Sun

Links : PDF - Abstract

Code :
Coursera

Keywords : style - adversarial - attacks - text - code -

Leave a Reply

Your email address will not be published. Required fields are marked *