Associate Professor, Carnegie Mellon University
3 papers at NeurIPS 2025
We prove that Private Evolution, a practical algorithm for differentially private synthetic data generation, converges in Wasserstein distance as the number of samples increases under a new, realistic theoretical model
We propose a sequential test for auditing differential privacy that detects violations with sample sizes that are orders of magnitude smaller than existing methods.
Two closed-form empirical Bernstein inequalities that are as sharp as the non-empirical ("oracle") Bernstein inequality for matrices.
