Search papers, labs, and topics across Lattice.
This paper establishes that the Gaussian mechanism is asymptotically optimal for differential privacy as the dimensionality of data approaches infinity, demonstrating that no additive-noise mechanism can surpass its privacy-utility tradeoff under strong privacy settings. Additionally, the authors introduce a new family of Spherical Generalized Gamma mechanisms that includes both the Gaussian and the recently proposed \(\ell_2\) mechanism, revealing specific instances that outperform these established methods in low-dimensional contexts. The findings also provide a resolution to a previous open question regarding the composition of the \(\ell_2\) mechanism, thereby advancing the understanding of differential privacy mechanisms in varying dimensional settings.
The Gaussian mechanism remains unbeatable in high dimensions, but new Spherical Generalized Gamma mechanisms reveal surprising advantages in low dimensions.
The additive noise mechanism is a foundational tool for differential privacy (DP) of $T$-dimensional real-valued vector queries. The Gaussian mechanism, utilizing Gaussian noise, is the mostly widely used such mechanism, due to its simplicity and strong privacy guarantees. In this work, we provide justification for this choice, showing that as the dimension $T\to\infty$, no additive-noise mechanism can asymptotically improve on the Gaussian mechanism's privacy--utility tradeoff for the strong privacy settings typically used.We also develop a new family of \emph{Spherical Generalized Gamma} DP mechanisms, which contains both the Gaussian mechanism and the recently studied $\ell_2$ mechanism (Joseph \emph{et al.}, ICML 2025). We identify members of this family that outperform both the Gaussian and $\ell_2$ mechanisms in certain low-dimensional settings, and show tight composition of all mechanisms in this family, answering an open question of Joseph \emph{et al.}~regarding the $\ell_2$ mechanism.
