English(EN) RPP: A Certified Poisoned-Sample Detection Framework for Backdoor Attacks under Dataset Imbalance

新框架RPP检测后门攻击中的中毒样本

作者 PulseAugur 编辑部 · [1 个来源] · 2026-10-12 04:00

研究人员开发了一个名为随机概率扰动（RPP）的新框架，用于检测深度神经网络后门攻击中使用的中毒样本。该框架即使在数据集不平衡的情况下也能有效工作，而数据集不平衡是现实场景中可能加剧后门漏洞的常见问题。RPP在黑盒设置下运行，仅依赖模型输出概率来识别被操纵的输入，并提供可检测性的可证明保证。 AI

影响这项研究可能带来更强大的防御措施，以应对利用现实世界人工智能应用中数据集不平衡的复杂攻击。

排序理由这是一篇详细介绍机器学习模型安全漏洞检测新框架的研究论文。[lever_c_demoted from research: ic=1 ai=1.0]

AI 生成摘要 · Google Gemini · 来自 1 个来源。我们如何撰写摘要 →

本文如何被排名

Signal score: 10 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 这是一篇详细介绍机器学习模型安全漏洞检测新框架的研究论文。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration: Single-source cluster; Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics: paper, safety; Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance: High; Clearly on-topic for AI-industry coverage.
Story freshness: Same-day; Cluster formed today. Ranking reflects the current source set at time of score.

完整方法见我们的编辑标准。

arXiv cs.LG TIER_1 English(EN) · Miao Lin, Feng Yu, Rui Ning, Lusi Li, Jiawei Chen, Qian Lou, Mengxin Zheng, Chunsheng Xin, Hongyi Wu · 2026-10-12 04:00

RPP：一种用于数据集不平衡下后门攻击的认证投毒样本检测框架

arXiv:2602.00183v2 Announce Type: replace-cross Abstract: Deep neural networks are highly susceptible to backdoor attacks, yet most defense methods to date rely on balanced data, overlooking the pervasive class imbalance in real-world scenarios that can amplify backdoor threats. …