Lalitha Sankar

GeoClip: Geometry-Aware Clipping for Differentially Private SGD

Optimizing $(L_0, L_1)$-Smooth Functions by Gradient Methods

Thumb on the Scale: Optimal Loss Weighting in Last Layer Retraining

CORAL: Disentangling Latent Representations in Long-Tailed Diffusion

Generalized Smooth Stochastic Variational Inequalities: Almost Sure Convergence and Convergence Rates

Optimizing Noise Distributions for Differential Privacy

Enhancing Robustness of Last Layer Two-Stage Fair Model Corrections