Enhancing the Transferability of Generative Targeted Adversarial Attacks via Cosine-Based Logit Alignment
Mathematics · 19.04.2026 · 14 (8) · 1370
DOI: 10.3390/math14081370 ↗Kurz gefasst
Fachartikel in der Zeitschrift Mathematics, online erschienen am 19.04.2026. Verfasst von Tengfei Shi und 2 weiteren Autor:innen. Die Arbeit nennt 52 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Adversarial examples reveal critical vulnerabilities in deep neural networks, posing significant risks in real-world deployment. In black-box settings, transferable targeted attacks rely on surrogate models but often suffer from low success rates. We argue that this limitation arises not only from surrogate-boundary overfitting but also from insufficient alignment with the target semantic space, which restricts the ability of adversarial examples to encode target-specific characteristics. To address this issue, we propose Cosine-Based Logit Alignment (CBLA), a unified framework for transferable targeted attacks. CBLA replaces the conventional cross-entropy loss with a cosine similarity objective to enhance directional alignment in logit space and alleviate gradient saturation. In addition, a semantic-invariant transformation strategy is introduced to improve structural consistency and cross-model generalization. Experiments on the ImageNet validation set demonstrate that CBLA consistently improves targeted attack success rates, achieving an average gain of 4.55% over strong baselines across multiple architectures.
Zitieren
APA
Shi, T., Wang, S., & Liu, B. (2026). Enhancing the Transferability of Generative Targeted Adversarial Attacks via Cosine-Based Logit Alignment. Mathematics, 14(8), 1370. https://doi.org/10.3390/math14081370
Vancouver
Shi T, Wang S, Liu B. Enhancing the Transferability of Generative Targeted Adversarial Attacks via Cosine-Based Logit Alignment. Mathematics. 2026;14(8):1370. doi:10.3390/math14081370
BibTeX
@article{Shi2026,
author = {Shi, Tengfei and Wang, Shihai and Liu, Bin},
title = {Enhancing the Transferability of Generative Targeted Adversarial Attacks via Cosine-Based Logit Alignment},
journal = {Mathematics},
year = {2026},
volume = {14},
number = {8},
pages = {1370},
doi = {10.3390/math14081370}
}
EUVIMED