English(EN) ShadowFPT: Backdooring Federated Prompt Tuning via Shadow Triggers

新的ShadowFPT攻击利用了联邦提示调优中冻结的主干网络

作者 PulseAugur 编辑部 · [1 个来源] · 2026-10-19 04:00

研究人员开发了一种名为ShadowFPT的新后门攻击，该攻击针对大型视觉语言模型中的联邦提示调优（FPT）。该攻击利用了模型的冻结主干网络，允许恶意客户端通过优化针对冻结编码器的“影子触发器”来注入后门行为。该触发器将触发的输入错误地引导至表示空间中的目标类别，而客户端的提示更新则保持隐蔽。ShadowFPT在保持干净准确率的同时显著提高了攻击成功率，表明仅关注提示更新异常的防御措施是不够的。 AI

影响突显了联邦学习在大型语言模型方面的新漏洞，可能需要新的防御机制。

排序理由详细介绍AI模型新攻击方法的学术论文。[lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv cs.AI 阅读 →

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: 9 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 详细介绍AI模型新攻击方法的学术论文。[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。

报道来源 [1]

arXiv cs.AI TIER_1 English(EN) · Kun Zhai, Teng Li, Yunhao Feng, Xingjun Ma · 2026-10-19 04:00

ShadowFPT：通过影子触发器对联邦提示调优进行后门攻击

arXiv:2610.22871v1 Announce Type: new Abstract: Federated Prompt Tuning (FPT) adapts large vision--language models by freezing the pretrained backbone and optimizing only lightweight prompt parameters across clients. Although this design improves communication and parameter effic…

本文如何被排名

报道来源 [1]

ShadowFPT：通过影子触发器对联邦提示调优进行后门攻击

相关实体

相关话题