paper-with-me

홈 › Papers

A Note on Implementation Errors in Recent Adaptive Attacks Against Multi-Resolution Self-Ensembles

2025-01-24 · Stanislav Fort

This note documents an implementation issue in recent adaptive attacks (Zhang et al. [2024]) against the multi-resolution self-ensemble defense (Fort and Lakshminarayanan [2024]). The implementation allowed adversarial perturbations to exceed the standard $L_\infty = 8/255$ bound by up to a factor of 20$\times$, reaching magnitudes of up to $L_\infty = 160/255$. When attacks are properly constrained within the intended bounds, the defense maintains non-trivial robustness. Beyond highlighting the importance of careful validation in adversarial machine learning research, our analysis reveals an intriguing finding: properly bounded adaptive attacks against strong multi-resolution self-ensembles often align with human perception, suggesting the need to reconsider how we measure adversarial robustness.

📄 PDF Abstract BibTeX arXiv:2501.14496

Code (0)

등록된 구현이 없습니다.

Tasks

Adversarial Robustness

Methods 이 논문이 사용한 방법론

ALIGN In the ALIGN method, visual and language representations are jointly trained from noisy image alt-text data. The image and text encoders are learned via contrastive loss…

Similar Papers 제목 키워드 기반

Adaptive Nearest Neighbor Machine Translation

2021-05-27 · ACL 2021 5 · Xin Zheng, Zhirui Zhang, Junliang Guo, ShuJian Huang 외

kNN-MT, recently proposed by Khandelwal et al. (2020a), successfully combines pre-trained neural machine translation (NMT) model with token-level k-nearest-neighbor (kNN) retrieval to improve the translation accuracy. Ho…

Machine TranslationNMTRetrievalTranslation

Retriever-and-Memory: Towards Adaptive Note-Enhanced Retrieval-Augmented Generation

2024-10-11 · Ruobing Wang, Daren Zha, Shi Yu, Qingfei Zhao 외

Retrieval-Augmented Generation (RAG) mitigates issues of the factual errors and hallucinated outputs generated by Large Language Models (LLMs) in open-domain question-answering tasks (OpenQA) via introducing external kno…

Open-Domain Question AnsweringQuestion AnsweringRAGRetrieval+1

Buffer Zone based Defense against Adversarial Examples in Image Classification

2021-01-01 · Kaleel Mahmood, Phuong Ha Nguyen, Lam M. Nguyen, Thanh V Nguyen 외

Recent defenses published at venues like NIPS, ICML, ICLR and CVPR are mainly focused on mitigating white-box attacks. These defenses do not properly consider adaptive adversaries. In this paper, we expand the scope of t…

Adversarial RobustnessClassificationGeneral Classificationimage-classification+1

Improved Robustness Against Adaptive Attacks With Ensembles and Error-Correcting Output Codes

2023-03-04 · Thomas Philippon, Christian Gagné

Neural network ensembles have been studied extensively in the context of adversarial robustness and most ensemble-based approaches remain vulnerable to adaptive attacks. In this paper, we investigate the robustness of Er…

Adversarial RobustnessDiversity

Generalization Error Matters in Decentralized Learning Under Byzantine Attacks

2024-07-11 · Haoxiang Ye, Qing Ling

Recently, decentralized learning has emerged as a popular peer-to-peer signal and information processing paradigm that enables model training across geographically distributed agents in a scalable manner, without the pre…