EMNLP Findings2023
Adversarial Text Generation by Search and Learning
Guoyi Li, Bingkang Shi, Zongzhen Liu, Dehan Kong, Yulei Wu, Xiaodan Zhang, Longtao Huang, Honglei Lyu
Abstract
Recent research has shown that evaluating the robustness of natural language processing models using textual attack methods is significant. However, most existing text attack methods only use heuristic replacement strategies or language models to generate replacement words at the word level. The blind pursuit of high attack success rates makes it difficult to ensure the quality of the generated adversarial text. As a result, adversarial text is often difficult for humans to understand. In fact, many methods that perform well in terms of text attacks often generate adversarial text with poor quality. To address this important gap, our work treats blackbox text attack as an unsupervised text generation problem and proposes a search and learning framework for Adversarial Text Generation by Search and Learning (ATGSL) and develops three adversarial attack methods (ATGSL-SA, ATGSL-BM, ATGSL-FUSION) for black-box text attacks. We first apply a heuristic search attack algorithm (ATGSL-SA) and a linguistic thesaurus to generate adversarial samples with high semantic similarity. After this process, we train a conditional generative model to learn from the search results while smoothing out search noise. Moreover, we design an efficient ATGSL-BM attack algorithm based on the text generator. Furthermore, we propose a hybrid attack method (ATGSL-FUSION) that integrates the advantages of ATGSL-SA and ATGSL-BM to enhance attack effectiveness. Our proposed attack algorithms are significantly superior to the most advanced methods in terms of attack efficiency and adversarial text quality.
Ask Lune
Know this paper in minutes, not an afternoon.
Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.
4 top-tier papers have cited it since. Know what they found before your reviewers do.
Lune can draw on
the full paper · the top-tier papers that cite it · related work
Free to start. No credit card required.
Lineage
- Are LLMs Stable Formal Logic Translators in Logical Reasoning Across Linguistically Diversified Texts? cites Is BERT Really Robust? A Strong Baseline for Natural Language Attack on Text Classification and Entailment
- Guided Perturbation Sensitivity (GPS): Detecting Adversarial Text via Embedding Stability and Word Importance cites Is BERT Really Robust? A Strong Baseline for Natural Language Attack on Text Classification and Entailment
- Query-Efficient and Scalable Black-Box Adversarial Attacks on Discrete Sequential Data via Bayesian Optimization cites Is BERT Really Robust? A Strong Baseline for Natural Language Attack on Text Classification and Entailment
- Unsupervised Text Generation by Learning from Search cites Iterative Edit-Based Unsupervised Sentence Simplification
Cited by top-tier papers4
- Distractor Generation in Multiple-Choice Tasks: A Survey of Methods, Datasets, and EvaluationElaf Alhazmi, Quan Sheng, Wei Emma Zhang, Munazza Zaib et al.EMNLP 2024 · 15 citations
- SECA: Semantically Equivalent and Coherent Attacks for Eliciting LLM HallucinationsBuyun Liang, Liangzu Peng, Jinqi Luo, Darshan Thaker et al.NeurIPS 2025 · 11 citations
- Are LLMs Stable Formal Logic Translators in Logical Reasoning Across Linguistically Diversified Texts?Qingchuan Li, Jiatong Li, Zirui Liu, Mingyue Cheng et al.WWW 2026 · 1 citation
- Guided Perturbation Sensitivity (GPS): Detecting Adversarial Text via Embedding Stability and Word ImportanceBryan E. Tuck, Rakesh M. VermaAAAI 2026 · 1 citation
Builds on6
- Is BERT Really Robust? A Strong Baseline for Natural Language Attack on Text Classification and EntailmentDi Jin, Zhijing Jin, Joey Tianyi Zhou, Peter SzolovitsAAAI 2020 · 1,333 citations
- Climbing towards NLU: On Meaning, Form, and Understanding in the Age of DataEmily M. Bender, Alexander KollerACL 2020 · 914 citations
- Unsupervised Text Generation by Learning from SearchJingjing Li, Zichao Li, Lili Mou, Xin Jiang et al.NeurIPS 2020 · 60 citations
- Iterative Edit-Based Unsupervised Sentence SimplificationDhruv Kumar, Lili Mou, Lukasz Golab, Olga VechtomovaACL 2020 · 58 citations
- Query-Efficient and Scalable Black-Box Adversarial Attacks on Discrete Sequential Data via Bayesian OptimizationDeokjae Lee, Seungyong Moon, Junhyeok Lee, Hyun Oh SongICML 2022 · 52 citations
Related papers
- HQA-Attack: Toward High Quality Black-Box Hard-Label Adversarial Attack on TextHan Liu, Zhi Xu, Xiaotong Zhang, Feng Zhang et al.NeurIPS 2023 · 32 citations
- Overcoming Black-box Attack Inefficiency with Hybrid and Dynamic Select AlgorithmsAbhinay Shankar Belde, Rohit Ramkumar, Jonathan RusertEMNLP Findings 2025 · 1 citation
- Bridge the Gap Between CV and NLP! A Gradient-based Textual Adversarial Attack FrameworkLifan Yuan, Yichi Zhang, Yangyi Chen, Wei WeiACL Findings 2023 · 27 citations
- Word-level Textual Adversarial Attacking as Combinatorial OptimizationYuan Zang, Fanchao Qi, Chenghao Yang, Zhiyuan Liu et al.ACL 2020 · 188 citations
- Experience Speaks Louder: Black-box Hard-label Adversarial Attack through Reinforcement LearningYilun Jin, Kun Zhu, Feng Tang, Jiaxuan Shi et al.KDD 2025 · 1 citation
