Devansh Gupta

guptadev (at) usc.edu Google Scholar View publications

About me

I am a fourth-year Computer Science Ph.D. student at the University of Southern California, where I am part of the CS Theory Group and the ODDS Group. I am advised by Meisam Razaviyayn and Vatsal Sharan.

I am broadly interested in fundamental problems in optimization and differential privacy.

Publications

Differentially Private Learning of Markov Chains: Near-Optimal Sample Complexity for Item and User Level Privacy

Devansh Gupta, Cristóbal Guzman, Enayat Ullah

ICML 2026 TPDP 2026

Efficient DP-SGD for LLMs with Randomized Clipping

Enayat Ullah, Sai Aparna Aketi, Devansh Gupta, Huanyu Zhang, Meisam Razaviyayn

TMLR 2026

Memory-Efficient Differentially Private Training with Gradient Random Projection

Alex Mulrooney, Devansh Gupta, James Flemings, Huanyu Zhang, Murali Annavaram, Meisam Razaviyayn, Xinwei Zhang

AISTATS 2025 OPT 2024 · Oral

Inherent Privacy of Zeroth-Order Projected Gradient Descent

Devansh Gupta, Meisam Razaviyayn, Vatsal Sharan

PPAI 2025 · AAAI

A Stochastic Optimization Framework for Private and Fair Learning From Decentralized Data

Devansh Gupta, A.S. Poornash, Andrew Lowy, Meisam Razaviyayn

ICLR 2023 TPDP 2023 AFCP 2022 · Oral

Stochastic Differentially Private and Fair Learning

Andrew Lowy, Devansh Gupta, Meisam Razaviyayn

ICASSP 2022

Time-Frequency and Geometric Analysis of Task-Dependent Learning in Raw Waveform Based Acoustic Models

Devansh Gupta, Vinayak Abrol