SOURCE-LINKED INTELLIGENCE

Revisiting the Provable-Auditable Privacy Gap of DP-SGD

arXiv · AI, language, vision and robotics · article · Aug 28, 2026 · UTC

Differential privacy (DP) has traditionally been used to provide theoretical upper bounds on an algorithm's stability to changing its training data. In modern private machine learning applications, achieving strong tradeoffs between utility and theoretical privacy is challenging, and thus one may optimistically hope that existing theoretical privacy analyses are loose. Recent work on privacy auditing has adopted a dual viewpoint, instead lower bounding the true privacy of an algorithm by constructing empirical distinguishing events. The auditing literature has thus far yielded a pessimistic ou

Read original source ↗ Open in workspace

recordType

paper

region

Global