论文标题
关于素数，日志损失分数和（否）隐私
On Primes, Log-Loss Scores and (No) Privacy
论文作者
论文摘要
会员推理攻击利用了将培训的客户数据训练模型的漏洞揭示到对手查询的漏洞。在最近提出的用于测量敏感数据集中隐私泄漏的审计工具的实施中，诸如日志损失分数之类的更精致的聚集物被暴露用于模拟推理攻击，并根据对手的预测来评估总隐私泄漏。在本文中，我们证明了此附加信息使对手能够在单个查询中完全准确地推断出任何数量的数据点的成员资格，从而造成完全的会员隐私漏洞。我们的方法避免了对对手的任何攻击模型培训或获得辅助知识的机会。此外，我们的算法对在攻击下的模型不可知，因此，即使对于不记住或过度拟合的模型，也可以使成员推理得出完美的推论。特别是，我们的观察结果提供了有关统计汇总信息泄漏程度以及如何利用它们的信息。
Membership Inference Attacks exploit the vulnerabilities of exposing models trained on customer data to queries by an adversary. In a recently proposed implementation of an auditing tool for measuring privacy leakage from sensitive datasets, more refined aggregates like the Log-Loss scores are exposed for simulating inference attacks as well as to assess the total privacy leakage based on the adversary's predictions. In this paper, we prove that this additional information enables the adversary to infer the membership of any number of datapoints with full accuracy in a single query, causing complete membership privacy breach. Our approach obviates any attack model training or access to side knowledge with the adversary. Moreover, our algorithms are agnostic to the model under attack and hence, enable perfect membership inference even for models that do not memorize or overfit. In particular, our observations provide insight into the extent of information leakage from statistical aggregates and how they can be exploited.
