Natalia Frumkin (Natasha)

PhD Candidate at UT Austin

I am a PhD candidate at the Univeristy of Texas at Austin with a focus on efficient deep learning and model compression. I'm lucky to work with my advisor, Diana Marculescu, and fellow colleagues in the EnyAC group.

News

Publications

Quamba2: A Robust and Scalable Post-training Quantization Framework for Selective State Space Models
Hung-Yueh Chiang, Chi-Chih Chiang, Natalia Frumkin, Kai-Chiang Wu, Mohammad S. Abdelfattah, Diana Marculescu
ICML, 2025

Quamba: A post-training quantization recipe for selective state space models
Hung-Yueh Chiang, Chi-Chih Chiang, Natalia Frumkin, Kai-Chiang Wu, Diana Marculescu
ICLR, 2024

Jumping through Local Minima: Quantization in the Loss Landscape of Vision Transformers
Natalia Frumkin, Dibakar Gope, Diana Marculescu
ICCV, 2023

MobileTL: On-device Transfer Learning with Inverted Residual Blocks
Hung-Yueh Chiang, Natalia Frumkin, Feng Liang, Diana Marculescu
AAAI, 2023

RL-Tune: A Deep Reinforcement Learning Assisted Layer-wise Fine-Tuning Approach for Transfer Learning
Tanvir Mahmud, Natalia Frumkin, Diana Marculescu
ICML Pre-Training Workshop, 2022

CNN-Based Indoor Occupant Localization via Active Scene Illumination
Jinyuan Zhao, Natalia Frumkin, Prakash Ishwar, Janusz Konrad
ICIP 2019

Privacy-Preserving Indoor Localization via Active Scene Illumination
Jinyuan Zhao, Natalia Frumkin, Janusz Konrad, Prakash Ishwar
CVPR Workshop, 2018

Work Experience

Meta

Research Scientist Intern

Summer 2023

Arm Research

Research Intern

Summer 2022

Advanced Micro Devices

Research Intern

Summer 2021

Los Alamos National Labs

Supercomputing Intern

Summer 2020

Red Hat

AI/Data Science Intern

Summer 2018

Education

UT Austin

ECE

PhD, graduation in December 2025 MSE, May 2023

Boston University

Computer Engineering

Bachelor of Science, May 2020