Model Releases
On Choosing the mu Parameter in Gaussian Differential Privacy
arXiv:2606.09582v1 Announce Type: new Abstract: Recent work argues for using Gaussian differential privacy (GDP) to report the privacy guarantees in privacy-preserving machine learning. We provide pri
arXiv:2606.09582v1 Announce Type: new 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 arepsilon to GDP mu 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 mu values across a useful range of parameters and recommend mu approx arepsilon/5 as a conservative general-purpose conversion.
Source: arXiv cs.LG | 2026-06-09