analiza aiqbench
This paper addresses critical vulnerabilities in federated learning by analyzing the transferability of adversarial examples across different client models, offering a robust defense mechanism grounded in adversarial training. The research provides theoretical insights into how data distribution affects model susceptibility to attacks like poisoning and Byzantine failures, validated through evaluations on real-life datasets. It represents a significant step forward in securing distributed machine learning systems against emerging threat vectors without relying on unrealistic assumptions about attacker ratios or client selection.
Rethinking the Transferable Adversarial Attacks and Robust Defense in Federated Learning
arXiv:2608.25133v1 Announce Type: new Abstract: The development of federated learning (FL) techniques has helped improve the privacy preservation of users' data and extended the applications of machine learning models. However, the involvement of a large number of users in FL also creates open opportunities for different adversaries, such as poisoning attacks, Byzantine attacks, and adversarial example attacks. Yet, recent research has disclosed that existing poisoning attacks and Byzantine attacks can not achieve satisfactory penetration in realistic FL scenarios caused by strong assumptions, \textit{e.g.,} client selection rate, and the ratio of malicious attackers. In this paper, the transferability of adversarial examples among different client models is analyzed to understand the relation between adversarial examples and clients' data distribution. Moreover, to mitigate the attacks of transferable adversarial examples, we design a defense mechanism stemming from the transferability of model robustness by adversarial training. As a result, through theoretical analysis of transferability, we gain insights into adversarial examples and the vulnerability of federated learning systems. Our proposed adversarial attack and defense methods are evaluated via real-life datasets in various settings to show their performance over the existing state-of-the-art methods.
Abstract:The development of federated learning (FL) techniques has helped improve the privacy preservation of users' data and extended the applications of machine learning models. However, the involvement of a large number of users in FL also creates open opportunities for different adversaries, such as poisoning attacks, Byzantine attacks, and adversarial example attacks. Yet, recent research has disclosed that existing poisoning attacks and Byzantine attacks can not achieve satisfactory penetration in realistic FL scenarios caused by strong assumptions, \textit{e.g.,} client selection rate, and the ratio of malicious attackers. In this paper, the transferability of adversarial examples among different client models is analyzed to understand the relation between adversarial examples and clients' data distribution. Moreover, to mitigate the attacks of transferable adversarial examples, we design a defense mechanism stemming from the transferability of model robustness by adversarial training. As a result, through theoretical analysis of transferability, we gain insights into adversarial examples and the vulnerability of federated learning systems. Our proposed adversarial attack and defense methods are evaluated via real-life datasets in various settings to show their performance over the existing state-of-the-art methods.
Machine Learning (cs.LG); Cryptography and Security (cs.CR)
arXiv:2608.25133 [cs.LG]
(or arXiv:2608.25133v1 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2608.25133 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 279 cuvinte
Subiecte conexe
Digital States 2026: AI Moves From Experiment to Infrastructure
govtech.com · 24 sept. 2026
Beyond Waveform Robustness: Robust Feature-Vocoder Adversarial Attacks on Automatic Speech Recognition
arXiv · 6 iun. 2026
ZeroGAR: Benchmarking the Adversarial Robustness of Zero-Shot Graph Models
arXiv · 30 sept. 2026
Reinforcement Learning Disrupts Gradient-Based Adversarial Optimization
arXiv · 12 iun. 2026