cs.LGJun 8, 2026

On Choosing the μμ Parameter in Gaussian Differential Privacy

Authors: Bogdan Kulynych, Antti Honkela

Organizations: Biomedical Data Science Center, Lausanne University Hospital · University of Helsinki

Abstract

Recent work argues for using Gaussian differential privacy (GDP) to report the privacy guarantees in privacy-preserving machine learning. We provide principled mappings from pure-DP ε\varepsilon to GDP μμ by matching the worst-case success of a strong-adversary membership inference attack in terms of three metrics: multiplicative advantage at fixed FPR, precision at fixed recall, and the standard privacy profile. We tabulate μμ values across a useful range of parameters and recommend μ≈ε/5μ\approx \varepsilon/5 as a conservative general-purpose conversion.

Explore similar work

CardsList
  1. Tight Auditing of Differential Privacy in MST and AIM

    Apr 20, 2026Georgi Ganev, Meenatchi Sundaram Muthu Selva Annamalai, Bogdan KulynychDifferential PrivacyAI Privacy Risks and Protection