Kunal Talwar

PREAMBLE: Private and Efficient Aggregation via Block Sparse Vectors

Instance-Optimality for Private KL Distribution Estimation

Adaptive Batch Size for Privately Finding Second-Order Stationary Points

Improved Sample Complexity for Private Nonsmooth Nonconvex Optimization

Enabling Differentially Private Federated Learning for Speech Recognition: Benchmarks, Adaptive Optimizers, and Gradient Clipping

Local Pan-privacy for Federated Analytics

Improved Sample Complexity for Private Nonsmooth Nonconvex Optimization

Faster Rates for Private Adversarial Bandits

Faster Rates for Private Adversarial Bandits

Private Online Learning via Lazy Algorithms

Private and Personalized Frequency Estimation in a Federated Setting

Instance-Optimal Private Density Estimation in the Wasserstein Distance

Federated Learning with Differential Privacy for End-to-End Speech Recognition