研究人员开发了Argus，一个旨在检测去中心化学习环境中后门攻击的新框架。该系统允许节点在没有中央服务器的情况下协同识别恶意的模型更新。Argus通过让节点共享潜在的触发器并利用结构相似性来区分真正的后门和由数据变化引起的误报。该框架还提供了理论收敛保证，并已证明在保持模型效用的同时显著降低了攻击成功率。 AI
影响 通过提供一种新颖的后门攻击防御方法，增强了协作式AI模型训练的安全性。
排序理由 该集群包含一篇详细介绍去中心化学习中后门检测新框架的学术论文。
AI 生成摘要 · Google Gemini · 来自 2 个来源。 我们如何撰写摘要 →
Decentralized learning (DL) is an emerging machine learning paradigm where nodes collaboratively train models without a central server. However, the collaborative nature of DL makes it vulnerable to backdoor attacks, where a model is taught to behave normally on standard inputs w…
Decentralized learning (DL) is an emerging machine learning paradigm where nodes collaboratively train models without a central server. However, the collaborative nature of DL makes it vulnerable to backdoor attacks, where a model is taught to behave normally on standard inputs w…