Lalitha Sankar — accepted AI-conference papers

2025

CORAL: Disentangling Latent Representations in Long-Tailed Diffusion

Esther Rodriguez, Monica Welfert, Samuel McDowell, Nathan Stromberg, Julian Antolin Camarena, Lalitha Sankar

NeurIPS 2025poster

Diffusion models have achieved impressive performance in generating high-quality and diverse synthetic data. However, their success typically assumes a class-balanced training distribution. In real-world settings, multi-class data often follow a long-tailed distribution, where standard diffusion mod…

Cited by 0 Source ↗Scholar ↗

2025

GeoClip: Geometry-Aware Clipping for Differentially Private SGD

Atefeh Gilani, Naima Tasnim, Lalitha Sankar, Oliver Kosut

NeurIPS 2025poster

Differentially private stochastic gradient descent (DP-SGD) is the most widely used method for training machine learning models with provable privacy guarantees. A key challenge in DP-SGD is setting the per-sample gradient clipping threshold, which significantly affects the trade-off between privacy…

Cited by 0 Source ↗Scholar ↗

2025

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

Daniil Vankov, Anton Rodomanov, Angelia Nedich, Lalitha Sankar, Sebastian U Stich

ICLR 2025poster

We study gradient methods for optimizing $(L_0, L_1)$-smooth functions, a class that generalizes Lipschitz-smooth functions and has gained attention for its relevance in machine learning. We provide new insights into the structure of this function class and develop a principled framework for analyzi…

Cited by 0 Source ↗PDF ↗Scholar ↗

2025

Optimizing Noise Distributions for Differential Privacy

Atefeh Gilani, Juan Felipe Gomez, Shahab Asoodeh, Flavio Calmon, Oliver Kosut, Lalitha Sankar

ICML 2025poster

We propose a unified optimization framework for designing continuous and discrete noise distributions that ensure differential privacy (DP) by minimizing Rényi DP, a variant of DP, under a cost constraint. Rényi DP has the advantage that by considering different values of the Rényi parameter $\alpha…

Cited by 0 Source ↗PDF ↗Scholar ↗

2025

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

Nathan Stromberg, Christos Thrampoulidis, Lalitha Sankar

NeurIPS 2025poster

While machine learning models become more capable in discriminative tasks at scale, their ability to overcome biases introduced by training data has come under increasing scrutiny. Previous results suggest that there are two extremes of parameterization with very different behaviors: the population…

Cited by 0 Source ↗Scholar ↗

2024

Enhancing Robustness of Last Layer Two-Stage Fair Model Corrections

Nathan Stromberg, Rohan Ayyagari, Sanmi Koyejo, Richard Nock, Lalitha Sankar

NeurIPS 2024poster

Last-layer retraining methods have emerged as an efficient framework for correcting existing base models. Within this framework, several methods have been proposed to deal with correcting models for subgroup fairness with and without group membership information. Importantly, prior work has demonstr…

Cited by 1 Source ↗PDF ↗Scholar ↗

2024

Generalized Smooth Variational Inequalities: Methods with Adaptive Stepsizes

Daniil Vankov, Angelia Nedich, Lalitha Sankar

ICML 2024poster

Variational Inequality (VI) problems have attracted great interest in the machine learning (ML) community due to their application in adversarial and multi-agent training. Despite its relevance in ML, the oft-used strong-monotonicity and Lipschitz continuity assumptions on VI problems are restrictiv…

Cited by 4 Source ↗PDF ↗Scholar ↗

2023

Smoothly Giving up: Robustness for Simple Models

Tyler Sypherd, Nathaniel Stromberg, Richard Nock, Visar Berisha, Lalitha Sankar

AISTATS 2023poster

There is a growing need for models that are interpretable and have reduced energy/computational cost (e.g., in health care analytics and federated learning). Examples of algorithms to train such models include logistic regression and boosting. However, one challenge facing these algorithms is that t…

Cited by 1 Source ↗PDF ↗Scholar ↗

2023

The Saddle-Point Method in Differential Privacy

Wael Alghamdi, Juan Felipe Gomez, Shahab Asoodeh, Flavio Calmon, Oliver Kosut, Lalitha Sankar

ICML 2023poster

We characterize the differential privacy guarantees of privacy mechanisms in the large-composition regime, i.e., when a privacy mechanism is sequentially applied a large number of times to sensitive data. Via exponentially tilting the privacy loss random variable, we derive a new formula for the pri…

Cited by 13 Source ↗PDF ↗Scholar ↗

2022

Being Properly Improper

Tyler Sypherd, Richard Nock, Lalitha Sankar

ICML 2022spotlight

Properness for supervised losses stipulates that the loss function shapes the learning algorithm towards the true posterior of the data generating distribution. Unfortunately, data in modern machine learning can be corrupted or twisted in many ways. Hence, optimizing a proper loss function on twiste…

Cited by 11 Source ↗PDF ↗Scholar ↗