English(EN) TwinMark: A Unified Watermark for Provable Survival Under Feature and Logit Distillation

新的水印方案TwinMark保护AI模型免受蒸馏攻击

作者 PulseAugur 编辑部 · [1 个来源] · 2026-09-17 04:00

研究人员开发了TwinMark，这是一种新颖的水印技术，旨在保护AI模型免受蒸馏攻击。该方法使用两个互补的线性函数，一个基于特征协方差，另一个基于类条件logit，来嵌入秘密水印。TwinMark旨在能够抵抗各种蒸馏方法，包括KL知识蒸馏和特征匹配蒸馏，并在多个数据集和模型架构上证明了其有效性。 AI

影响这项水印技术可以增强AI模型在未经授权的复制和操纵方面的安全性和可追溯性。

排序理由该条目是一篇学术论文，详细介绍了一种用于AI模型水印的新技术方法。[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: 11 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 该条目是一篇学术论文，详细介绍了一种用于AI模型水印的新技术方法。[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.LG TIER_1 English(EN) · Redwanul Karim, Tobias Feigl, Christopher Mutschler, Felix Ott · 2026-09-17 04:00

TwinMark：一种可证明在特征和logit蒸馏下生存的统一水印

arXiv:2609.19011v1 Announce Type: new Abstract: We propose TwinMark, a watermarking scheme that reads a single SHAKE128 secret through two complementary linear functionals of model-output summaries: a covariance projector against the carrier-set covariance (cov-Feat) and a class-…

本文如何被排名

报道来源 [1]

TwinMark：一种可证明在特征和logit蒸馏下生存的统一水印

相关实体

相关话题