Word Level Adversarial

Word-level adversarial attacks aim to subtly alter text inputs to fool natural language processing (NLP) models into making incorrect predictions, highlighting vulnerabilities in these systems. Current research focuses on developing more effective attack methods, such as those leveraging large language models to generate natural-sounding adversarial examples, and designing robust defenses, including techniques that randomize latent representations or utilize local explainability methods. This research is crucial for improving the reliability and security of NLP applications, as it directly addresses the susceptibility of these models to malicious manipulation.

Papers

November 12, 2024

Deceiving Question-Answering Models: A Hybrid Word-Level Adversarial Approach
Jiyao Li, Mingze Ni, Yongshun Gong, Wei Liu
Adversarial Attack Full Model Constructive Approach QuEry Based Attack Word Level Adversarial

November 20, 2023

Generating Valid and Natural Adversarial Examples with Large Language Models
Zimu Wang, Wei Wang, Qi Chen, Qiufeng Wang, Anh Nguyen
Adversarial Attack Adversarial Example Word Level Adversarial

October 2, 2023

Fooling the Textual Fooler via Randomizing Latent Representations
Duy C. Hoang, Quang H. Nguyen, Saurav Manchanda, MinLong Peng, Kok-Seng Wong, Khoa D. Doan
Adversarial Attack Adversarial Example Adversarial Perturbation Latent Representation Word Level Adversarial Adversarial Text Perturbation

August 1, 2023

LimeAttack: Local Explainable Method for Textual Hard-Label Adversarial Attack
Hai Zhu, Zhaoqing Yang, Weiwei Shang, Yuren Wu
Adversarial Example Local Explanation Hard Label Attack Word Level Adversarial

July 31, 2023

Text-CRS: A Generalized Certified Robustness Framework against Textual Adversarial Attacks
Xinyu Zhang, Hanbin Hong, Yuan Hong, Peng Huang, Binghui Wang, Zhongjie Ba, Kui Ren
Textual Adversarial Attack Certified Robustness Synonym Substitution Attack Word Level Adversarial Insertion Based

May 24, 2023

How do humans perceive adversarial text? A reality check on the validity and naturalness of word-based adversarial attacks
Salijona Dyrmishi, Salah Ghamizi, Maxime Cordy
Adversarial Attack Adversarial Example Adversarial Perturbation Adversarial Text Concurrent Validity Visual Naturalness Text Attack Reality Check Word Level Adversarial

November 12, 2022

Generating Textual Adversaries with Minimal Perturbation
Xingyi Zhao, Lu Zhang, Depeng Xu, Shuhan Yuan
Adversarial Text Small Perturbation Attack Algorithm Word Level Adversarial

February 28, 2022

Robust Textual Embedding against Word-level Adversarial Attacks
Yichen Yang, Xiaosen Wang, Kun He
Language Model Robust Version Natural Language Processing Model Metric Learning Word Level Adversarial