PulseAugur
实时 16:34:51
English(EN) Differentially Private Datastore Generation for Retrieval-Augmented Inference

新方法为AI推理生成私有数据存储

研究人员开发了一个新的差分隐私数据存储生成框架,这对于使用检索增强推理的AI系统至关重要。他们基于哈希的方法对数据进行分区并添加校准噪声,以确保个人隐私同时保持效用。实验表明,在epsilon=5时,准确率仅下降2.6%,并且该方法显著降低了成员推理攻击的有效性。 AI

影响 能够安全地发布用于AI的数据存储,可能增加对设备上AI系统的信任和采用。

排序理由 这是一篇详细介绍差分隐私数据存储生成新方法的学术论文。

在 arXiv cs.IR (Information Retrieval) 阅读 →

AI 生成摘要 · Google Gemini · 来自 2 个来源。 我们如何撰写摘要 →

报道来源 [2]

  1. arXiv cs.LG TIER_1 English(EN) · Abdelrahman Abouelenein, Marwan Torki ·

    Differentially Private Datastore Generation for Retrieval-Augmented Inference

    arXiv:2606.01413v1 Announce Type: cross Abstract: It is crucial for modern on-device AI systems that rely on retrieval-augmented inference to release and share datastores without compromising individual privacy. This can be achieved using Differential Privacy (DP), which provides…

  2. arXiv cs.IR (Information Retrieval) TIER_1 English(EN) · Marwan Torki ·

    Differentially Private Datastore Generation for Retrieval-Augmented Inference

    It is crucial for modern on-device AI systems that rely on retrieval-augmented inference to release and share datastores without compromising individual privacy. This can be achieved using Differential Privacy (DP), which provides a formal guarantee that ensures individual contri…