PulseAugur
中
实时 09:34:14
English(EN) What Gradients Add to Text Leakage in Split Language Models, Counted per Token and per Document

研究发现分体式语言模型通过梯度泄露客户文本

一篇新的研究论文探讨了分体式语言模型相关的隐私风险,在这种模型中,客户在服务器上训练模型,而无需发送其原始文本。研究表明,能够访问模型公开权重观察者可以从训练过程中交换的激活和梯度中重建客户文本的很大一部分。具体来说,在GPT-2上的实验表明,添加梯度将 token 恢复率从 94.20% 提高到 97.38%,将精确文档恢复率从 13.71% 提高到 37.77%。该研究主张在 token 和文档级别报告泄露情况,并将分体式模型中传输的数据视为高度敏感。 AI

影响 强调了分体式学习设置中的重大隐私风险,可能影响敏感数据在分布式人工智能训练中的处理方式。

排序理由 学术论文,详细介绍了关于模型隐私的新发现。[lever_c_demoted from research: ic=1 ai=1.0]

在 Hugging Face Daily Papers 阅读 →

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

研究发现分体式语言模型通过梯度泄露客户文本

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
学术论文,详细介绍了关于模型隐私的新发现。[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
6 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]

  1. Hugging Face Daily Papers TIER_1 English(EN) ·

    梯度在拆分语言模型中对文本泄露的贡献,按 token 和文档计数

    Split learning lets a client train a language model on a server without sending its text. The client runs the first layers itself and sends the server only their output, a vector of numbers for each token. During training, the server sends gradients back. We show that an observer…