Sayantan Choudhury

Research

My research focuses on optimization theory for machine learning, with interests in extragradient methods, adaptive optimization, federated learning, min-max problems, and scalable methods for structured non-convex settings. I have also worked on conformal prediction and generative models.

Publications

1. One-shot Conditional Sampling: MMD meets Nearest Neighbors

Anirban Chatterjee, Sayantan Choudhury, and Rohan Hore

Accepted at ICML 2026.

Paper

2. Multiplayer Federated Learning: Reaching Equilibrium with Less Communication

TaeHo Yoon, Sayantan Choudhury, and Nicolas Loizou

Accepted at NeurIPS 2025.

Paper

3. Extragradient Method for (L0, L1)-Lipschitz Root-finding Problems

Sayantan Choudhury and Nicolas Loizou

Accepted at NeurIPS 2025.

Paper

4. Methods for Convex (L0, L1)-Smooth Optimization: Clipping, Acceleration, and Adaptivity

Eduard Gorbunov, Nazarii Tupitsa, Sayantan Choudhury, Alen Aliev, Peter Richtarik, Samuel Horvath, and Martin Takac

Accepted at ICLR 2025.

Paper

5. Remove that Square Root: Efficient Scale-Invariant Version of AdaGrad

Sayantan Choudhury, Nazarii Tupitsa, Nicolas Loizou, Samuel Horvath, Martin Takac, and Eduard Gorbunov

Accepted at NeurIPS 2024.

Paper

6. Communication-Efficient Gradient Descent-Accent Methods for Distributed Variational Inequalities: Unified Analysis and Local Updates

Siqi Zhang*, Sayantan Choudhury*, Sebastian Stich, and Nicolas Loizou

Accepted at ICLR 2024. * Equal contribution.

Paper

7. Single-Call Stochastic Extragradient Methods for Structured Non-monotone Variational Inequalities: Improved Analysis under Weaker Conditions

Sayantan Choudhury, Eduard Gorbunov, and Nicolas Loizou

Accepted at NeurIPS 2023.

Paper

Preprints

1. Multi-source Conformal Prediction: Leveraging Heterogeneity via Localization

Rohan Hore, Anirban Chatterjee, and Sayantan Choudhury

Preprint, 2026.

Paper

2. Gradient Clipping Beyond Vector Norms: A Spectral Approach for Matrix-Valued Parameters

Alexander Yukhimchuk, Mladen Kolar, Martin Takac, and Sayantan Choudhury

Preprint, 2026.

Paper

3. Muon with Nesterov Momentum: Heavy-Tailed Noise and (Randomized) Inexact Polar Decomposition

Sayantan Choudhury, Xiaoran Cheng, Martin Takac, Sen Na, and Mladen Kolar

Preprint, 2026.

Paper