Exploiting Class Probabilities for Black-box Sentence-level Attacks
Not on the map. Why?
Abstract (arXiv)
Sentence-level attacks craft adversarial sentences that are synonymous with correctly-classified sentences but are misclassified by the text classifiers. Under the black-box setting, classifiers are only accessible through their feedback to queried inputs, which is predominately available in the form of class probabilities. Even though utilizing class probabilities results in stronger attacks, due to the challenges of using them for sentence-level attacks, existing attacks use either no feedback or only the class labels. Overcoming the challenges, we develop a novel algorithm that uses class probabilities for black-box sentence-level attacks, investigate the effectiveness of using class probabilities on the attack's success, and examine the question if it is worthy or practical to use class probabilities by black-box sentence-level attacks. We conduct extensive evaluations of our attack comparing with the baselines across various classifiers and benchmark datasets.
Key facts. Cited by 1 paper in the map (data through 7 Oct 2026). Builds on 29 of its 42 references that are in the map. First posted on arXiv 5 Feb 2024.
Builds on
29 of 42 references.
References · 29
- 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,204
- 2019Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformercited by 14,491
- 2019RoBERTa: A Robustly Optimized BERT Pretraining Approachcited by 13,266
- 2022Generative Adversarial Networkscited by 5,603
- 2019BERTScore: Evaluating Text Generation with BERTcited by 5,462
- 2016Towards Evaluating the Robustness of Neural Networkscited by 2,976
- 2019XLNet: Generalized Autoregressive Pretraining for Language Understandingcited by 2,801
- 2015Character-level Convolutional Networks for Text Classificationcited by 2,293
- 2019Is BERT Really Robust? A Strong Baseline for Natural Language Attack on Text Classification and Entailmentcited by 663
- 2020BERT-ATTACK: Adversarial Attack Against BERT Using BERTcited by 488
- 2018Universal Sentence Encodercited by 485
- 2018Generating Natural Language Adversarial Examplescited by 422
17 more
- 2020BAE: BERT-based Adversarial Examples for Text Classificationcited by 316
- 2018Adversarial Example Generation with Syntactically Controlled Paraphrase Networkscited by 282
- 2021Gradient-based Adversarial Attacks against Text Transformerscited by 233
- 2020Contextualized Perturbation for Textual Adversarial Attackcited by 136
- 2020Optimus: Organizing Sentences via Pre-trained Modeling of a Latent Spacecited by 114
- 2020OpenAttack: An Open-source Textual Adversarial Attack Toolkitcited by 84
- 2018GenAttack: Practical Black-box Attacks with Gradient-Free Optimizationcited by 82
- 2021Generating Syntactically Controlled Paraphrases without Using Annotated Parallel Pairscited by 45
- 2021Grey-box Adversarial Attack And Defence For Sentiment Classificationcited by 28
- 2021Bridge the Gap Between CV and NLP! A Gradient-based Textual Adversarial Attack Frameworkcited by 27
- 2020MALCOM: Generating Malicious Comments to Attack Neural Fake News Detection Modelscited by 25
- 2022Query-Efficient and Scalable Black-Box Adversarial Attacks on Discrete Sequential Data via Bayesian Optimizationcited by 22
- 2021Multi-granularity Textual Adversarial Attack with Behavior Cloningcited by 21
- 2021A Strong Baseline for Query Efficient Attacks in a Black Box Settingcited by 16
- 2021Adversarial Stylometry in the Wild: Transferable Lexical Substitution Attacks on Author Profilingcited by 12
- 2022TextHacker: Learning based Hybrid Local Search Algorithm for Text Hard-label Adversarial Attackcited by 10
- 2021Adv-OLM: Generating Textual Adversaries via OLMcited by 2
Built on by
Citing papers per year: 2025: 1.
1 paper.