Optimal Differentially Private Sampling of Unbounded Gaussians
arXiv:2503.01766v11 citationsh-index: 4COLT
Originality Highly original
AI Analysis
This work addresses a problem significant for researchers and practitioners in the field of differential privacy, providing an incremental yet substantial improvement over previous results.
The authors tackled the problem of sampling from unbounded Gaussian distributions under differential privacy constraints, achieving a quadratic improvement with a sample complexity of O(d).
We provide the first $\widetilde{\mathcal{O}}\left(d\right)$-sample algorithm for sampling from unbounded Gaussian distributions under the constraint of $\left(\varepsilon, δ\right)$-differential privacy. This is a quadratic improvement over previous results for the same problem, settling an open question of Ghazi, Hu, Kumar, and Manurangsi.