English(EN) BiShield-TEE: On the (In-)Security of Unilateral Weight Obfuscation in On-Device TEE-Shielded LLM Partition

新的 BiShield-TEE 框架解决了设备端 LLM 的安全漏洞

作者 PulseAugur 编辑部 · [1 个来源] · 2026-10-16 04:00

研究人员发现了一种在依赖单边权重混淆的设备端大模型（LLM）部署方法中的漏洞。这种被称为解决方案确定性属性（SDP）的漏洞允许攻击者唯一确定混淆密钥，从而危及模型安全。为此，提出了一个名为 BiShield-TEE 的新框架，该框架采用输入和输出维度的双边混淆来创建模糊的解决方案，从而在多个有效候选者中隐藏真实密钥。实验表明，BiShield-TEE 能有效地将攻击准确率限制在基线水平，同时提高推理效率。 AI

影响增强了设备端 LLM 部署的安全性，通过解决知识产权和逆向工程问题，可能促使其更广泛的应用。

排序理由详细介绍 LLM 新安全框架的学术论文。[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: 17 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 详细介绍 LLM 新安全框架的学术论文。[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: Breaking (< 6h); Fresh story with cross-source coverage still developing. Ranking may shift as more sources report.

完整方法见我们的编辑标准。

报道来源 [1]

arXiv cs.AI TIER_1 English(EN) · Zhiyuan Wang, Yuezhi Zhou, Ziqi Zhang, Zhu Zhang, Zhixing Tan, Yongheng Deng · 2026-10-16 04:00

BiShield-TEE：设备端TEE保护的大模型分区中单方面权重混淆的（不）安全性

arXiv:2610.21565v1 Announce Type: cross Abstract: On-device deployment of large language models (LLMs) is increasingly adopted due to latency and privacy requirements, but exposes proprietary models to reverse engineering and intellectual property theft. Trusted Execution Environ…

本文如何被排名

报道来源 [1]

BiShield-TEE：设备端TEE保护的大模型分区中单方面权重混淆的（不）安全性

相关实体

相关话题