English(EN) Leaky Language Models: Stealing Architecture and Inference Optimizations via Per-Token Timing

新的LeakyLMs攻击通过时间戳揭示Gemini Flash 2.5架构

作者 PulseAugur 编辑部 · [1 个来源] · 2026-07-24 04:00

研究人员开发了一套名为LeakyLMs的新攻击方法，通过分析令牌生成的时间戳，可以推断出有关语言模型的专有信息，包括其架构和推理优化。此方法甚至可以通过远程API提取详细信息。一项攻击成功识别出Google Gemini Flash 2.5使用了具有约128K令牌草稿上下文窗口的推测性解码。另一项攻击可以恢复架构属性，如Transformer层数、隐藏层维度大小和注意力头数，在识别Llama模型配置方面准确率超过90%。 AI

影响揭示了已部署LLM的潜在安全漏洞，影响模型的安全性和架构保护方式。

排序理由详细介绍语言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.LG 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: 0 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 详细介绍语言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: 84 days old; Aged out of breaking-news scoring windows; ranking reflects the durable signal from the full source set.

完整方法见我们的编辑标准。

报道来源 [1]

arXiv cs.LG TIER_1 English(EN) · Sadegh Majidi, Niloofar Mireshghallah, Kazem Taram · 2026-07-24 04:00

泄露的语言模型：通过每令牌计时窃取架构和推理优化

arXiv:2607.20723v1 Announce Type: cross Abstract: This work presents LeakyLMs, a set of attacks that leak proprietary model, architecture, and deployment information from production language models. LeakyLMs is the first to demonstrate that key model and deployment details can be…

本文如何被排名

报道来源 [1]

泄露的语言模型：通过每令牌计时窃取架构和推理优化

相关实体

相关话题