mlsanityData to 7 Oct 2026

Contextualized Perturbation for Textual Adversarial Attack

arXiv 2009.07502 ↗First posted 16 Sep 2020cs.CLcited by 136

Not on the map. Why?

Abstract (arXiv)

Adversarial examples expose the vulnerabilities of natural language processing (NLP) models, and can be used to evaluate and improve their robustness. Existing techniques of generating such examples are typically driven by local heuristic rules that are agnostic to the context, often resulting in unnatural and ungrammatical outputs. This paper presents CLARE, a ContextuaLized AdversaRial Example generation model that produces fluent and grammatical outputs through a mask-then-infill procedure. CLARE builds on a pre-trained masked language model and modifies the inputs in a context-aware manner. We propose three contextualized perturbations, Replace, Insert and Merge, allowing for generating outputs of varied lengths. With a richer range of available strategies, CLARE is able to attack a victim model more efficiently with fewer edits. Extensive experiments and human evaluation demonstrate that CLARE outperforms the baselines in terms of attack success rate, textual similarity, fluency and grammaticality.

Key facts. Cited by 136 papers in the map (data through 7 Oct 2026). First posted on arXiv 16 Sep 2020.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 23; 2022: 25; 2023: 42; 2024: 28; 2025: 10; 2026: 8.

136 papers.

Sort:

By area:

  1. 2021Recent Advances in Natural Language Processing via Large Pre-Trained Language Models: A Surveycited by 457
  2. 2023Explainability for Large Language Models: A Surveycited by 331
  3. 2023From Pretraining Data to Language Models to Downstream Tasks: Tracking the Trails of Political Biases Leading to Unfair NLP Modelscited by 257
  4. 2021Gradient-based Adversarial Attacks against Text Transformerscited by 233
  5. 2021Polyjuice: Generating Counterfactuals for Explaining, Evaluating, and Improving Modelscited by 219
  6. 2021Measure and Improve Robustness in NLP Models: A Surveycited by 107
  7. 2022Why Should Adversarial Perturbations be Imperceptible? Rethink the Research Paradigm in Adversarial NLPcited by 99
  8. 2023MGTBench: Benchmarking Machine-Generated Text Detectioncited by 90
  9. 2021Towards Improving Adversarial Training of NLP Modelscited by 79
  10. 2023Robust Multi-bit Natural Language Watermarking through Invariant Featurescited by 76
  11. 2021NL-Augmenter: A Framework for Task-Sensitive Natural Language Augmentationcited by 72
  12. 2023VLATTACK: Multimodal Adversarial Attacks on Vision-Language Tasks via Pre-trained Modelscited by 66
38 more
  1. 2021Triggerless Backdoor Attack for NLP Tasks with Clean Labelscited by 61
  2. 2022BITE: Textual Backdoor Attacks with Iterative Trigger Injectioncited by 47
  3. 2022SemAttack: Natural Textual Attacks via Different Semantic Spacescited by 42
  4. 2024Changing Answer Order Can Decrease MMLU Accuracycited by 41
  5. 2023Did You Train on My Dataset? Towards Public Dataset Protection with Clean-Label Backdoor Watermarkingcited by 37
  6. 2023Smaller Language Models are Better Black-box Machine-Generated Text Detectorscited by 35
  7. 2021FlipDA: Effective and Robust Data Augmentation for Few-Shot Learningcited by 32
  8. 2023Text-CRS: A Generalized Certified Robustness Framework against Textual Adversarial Attackscited by 32
  9. 2021Perturbing Inputs for Fragile Interpretations in Deep Natural Language Processingcited by 30
  10. 2023Evaluating the Robustness of Text-to-image Diffusion Models against Real-world Attackscited by 24
  11. 2021Achieving Model Robustness through Discrete Adversarial Trainingcited by 23
  12. 2022Generalized but not Robust? Comparing the Effects of Data Modification Methods on Out-of-Domain Generalization and Adversarial Robustnesscited by 23
  13. 2022Phrase-level Textual Adversarial Attack with Label Preservationcited by 23
  14. 2024Data Contamination Calibration for Black-box LLMscited by 23
  15. 2022TextGrad: Advancing Robustness Evaluation in NLP by Gradient-Driven Optimizationcited by 21
  16. 2023How do humans perceive adversarial text? A reality check on the validity and naturalness of word-based adversarial attackscited by 18
  17. 2021A Strong Baseline for Query Efficient Attacks in a Black Box Settingcited by 16
  18. 2021Text Counterfactuals via Latent Optimization and Shapley-Guided Searchcited by 16
  19. 2022BERT Rankers are Brittle: a Study using Adversarial Document Perturbationscited by 16
  20. 2024HQA-Attack: Toward High Quality Black-Box Hard-Label Adversarial Attack on Textcited by 16
  21. 2022AEON: A Method for Automatic Evaluation of NLP Test Casescited by 15
  22. 2022TASA: Deceiving Question Answering Models by Twin Answer Sentences Attackcited by 15
  23. 2024Advancing the Robustness of Large Language Models through Self-Denoised Smoothingcited by 15
  24. 2024LLM Whisperer: An Inconspicuous Attack to Bias LLM Responsescited by 15
  25. 2021Accurate, yet inconsistent? Consistency Analysis on Language Understanding Modelscited by 13
  26. 2023Exploring Multimodal Approaches for Alzheimer's Disease Detection Using Patient Speech Transcript and Audio Datacited by 13
  27. 2021Token-Modification Adversarial Attacks for Natural Language Processing: A Surveycited by 12
  28. 2022Adversarial Training for Improving Model Robustness? Look at Both Prediction and Interpretationcited by 12
  29. 2021Unifying Model Explainability and Robustness for Joint Text Classification and Rationale Extractioncited by 11
  30. 2023Two-in-One: A Model Hijacking Attack Against Text Generation Modelscited by 11
  31. 2023Toward Adversarial Training on Contextualized Language Representationcited by 10
  32. 2024VQAttack: Transferable Adversarial Attacks on Visual Question Answering via Pre-trained Modelscited by 10
  33. 2021Consistency Training with Virtual Adversarial Discrete Perturbationcited by 9
  34. 2021Semantically Distributed Robust Optimization for Vision-and-Language Inferencecited by 9
  35. 2023LimeAttack: Local Explainable Method for Textual Hard-Label Adversarial Attackcited by 9
  36. 2024A Modified Word Saliency-Based Adversarial Attack on Text Classification Modelscited by 9
  37. 2023TextShield: Beyond Successfully Detecting Adversarial Sentences in Text Classificationcited by 8
  38. 2024Toward Robust RALMs: Revealing the Impact of Imperfect Retrieval on Retrieval-Augmented Language Modelscited by 8

50 most cited of 136 All 136 (JSON)