paper-with-me

Papers

Revisiting Character-level Adversarial Attacks for Language Models

2024-05-07 · Elias Abad Rocamora, Yongtao Wu, Fanghui Liu, Grigorios G. Chrysos, Volkan Cevher

Adversarial attacks in Natural Language Processing apply perturbations in the character or token levels. Token-level attacks, gaining prominence for their use of gradient-based methods, are susceptible to altering sentence semantics, leading to invalid adversarial examples. While character-level attacks easily maintain semantics, they have received less attention as they cannot easily adopt popular gradient-based methods, and are thought to be easy to defend. Challenging these beliefs, we introduce Charmer, an efficient query-based adversarial attack capable of achieving high attack success rate (ASR) while generating highly similar adversarial examples. Our method successfully targets both small (BERT) and large (Llama 2) models. Specifically, on BERT with SST-2, Charmer improves the ASR in 4.84% points and the USE similarity in 8% points with respect to the previous art. Our implementation is available in https://github.com/LIONS-EPFL/Charmer.

📄 PDF Abstract BibTeX arXiv:2405.04346

Code (1)

lions-epfl/charmer 공식 구현 pytorch

Tasks

Adversarial AttackSentenceSST-2

Methods 이 논문이 사용한 방법론

Refunds@Expedia|||How do I get a full refund from Expedia? “How do I get a full refund from Expedia? How do I get a full refund from Expedia? – Call ☎️ +1-(888) 829 (0881) or +1-805-330-4056 or +1-805-330-4056 for Quick Help &…
Attention 설명 없음
Weight Decay 설명 없음
Attention Dropout Attention Dropout is a type of dropout used in attention-based architectures, where elements are randomly dropped out of the…
Dropout Dropout is a regularization technique for neural networks that drops a unit (along with connections) at training time with a specified probability $p$ (a common value is…
Residual Connection 설명 없음
Softmax The Softmax output function transforms a previous layer's output into a vector of probabilities. It is commonly used for multiclass classification. Given an input vector $x$…
WordPiece 설명 없음

Similar Papers 제목 키워드 기반

Revisiting the Adversarial Robustness of Vision Language Models: a Multimodal Perspective

2024-04-30 · Wanqi Zhou, Shuanghao Bai, Danilo P. Mandic, Qibin Zhao 외

Pretrained vision-language models (VLMs) like CLIP exhibit exceptional generalization across diverse downstream tasks. While recent studies reveal their vulnerability to adversarial attacks, research to date has primaril…

Adversarial DefenseAdversarial RobustnessAdversarial Text

Adversarial Text Normalization

2022-06-08 · NAACL (ACL) 2022 7 · Joanna Bitton, Maya Pavlova, Ivan Evtimov

Text-based adversarial attacks are becoming more commonplace and accessible to general internet users. As these attacks proliferate, the need to address the gap in model robustness becomes imminent. While retraining on a…

Adversarial TextNatural Language InferenceText Normalization

Robustness of Large Language Models Against Adversarial Attacks

2024-12-22 · Yiyi Tao, Yixian Shen, Hang Zhang, Yanxin Shen 외

The increasing deployment of Large Language Models (LLMs) in various applications necessitates a rigorous evaluation of their robustness against adversarial attacks. In this paper, we present a comprehensive study on the…

Sentiment AnalysisSentiment ClassificationSST-2

Adversarial Evasion Attack Efficiency against Large Language Models

2024-06-12 · João Vitorino, Eva Maia, Isabel Praça

Large Language Models (LLMs) are valuable for text classification, but their vulnerabilities must not be disregarded. They lack robustness against adversarial examples, so it is pertinent to understand the impacts of dif…

Adversarial DefenseClassificationSentiment AnalysisSentiment Classification+2

Revisiting Ensembles in an Adversarial Context: Improving Natural Accuracy

2020-02-26 · Aditya Saligrama, Guillaume Leclerc

A necessary characteristic for the deployment of deep learning models in real world applications is resistance to small adversarial perturbations while maintaining accuracy on non-malicious inputs. While robust training …