About Me
I’m a second-year PhD student at MIT CSAIL, advised by Prof. Sara Beery! I am grateful to be supported by the MIT Jameel Clinic Fellowship and the NSF GRFP.
Previously, I was a master’s student at MIT CSAIL in the HealthyML Lab advised by Prof. Marzyeh Ghassemi, where I worked on detecting & mitigating distribution shift. Last summer, I was a visiting student at ETH Zurich advised by Prof. Fanny Yang, where I worked on optimal transport and differential privacy. I also attended MIT for undergrad and double majored in computer science and mathematics (Course 6 & 18) with a concentration in Ancient and Medieval Studies. I’ve had the pleasure to intern at Microsoft Research (x3) and Apple Research.
I’m also on Twitter and Google Scholar. Feel free to contact me at nhulkund@mit.edu!
Research Interests
I’m interested in building reliable machine learning systems, balancing tradeoffs of performance, efficiency, privacy, and robustness. Some current questions I am thinking about:
- How do we leverage knowledge from large-scale foundation models to be applied for specialized downstream tasks?
- How do we select data for large models carefully for increased robustness and resource efficiency?
- Detecting distribution shifts is well studied on datasets such as ImageNet, how well does this translate to other domains?
Working Papers
- Challenges and Pitfalls of Geopolitical Data Sharing Policy for Ecological Data
Neha Hulkund, Millie Chapman, Ruth Oliver, Sara Beery
Under Review, presented at CVPR 2024 Responsible Data Workshop
Publications
-
Privacy-preserving data release leveraging optimal transport and particle gradient descent
Konstantin Donhauser, Javier Abad, Neha Hulkund, Fanny Yang
ICML 2024, ICLR 2024 Privacy Regulation and Protection in ML Workshop
arxiv -
Predicting Out-of-Domain Generalization with Local Manifold Smoothness
Nathan Ng, Neha Hulkund, Kyunghyun Cho, Marzyeh Ghassemi
TMLR
arxiv -
Interpretable Distribution Shift Detection using Optimal Transport
Neha Hulkund, Nicolo Fusi, Jennifer Wortman Vaughan, David Alvarez-Melis
ICML 2022 DataPerf Workshop
arxiv -
GAN-based Data Augmentation for Chest X-ray Classification
Shobhita Sundaram*, Neha Hulkund* (equal contribution)
KDD 2021 DSHealth Workshop Spotlight Presentation
arxiv -
The Limits of Algorithmic Stability for Robustness to Distribution Shift
Neha Hulkund, Vinith Suriyakumar, Taylor Killian, Marzyeh Ghassemi
NeurIPS 2022 Women in Machine Learning Workshop
pdf poster
Class Projects
-
Facilitating Fairness through Distributionally Robust Finetuning
Final Project for MIT class 6.864: Natural Language Processing
pdf -
Extension of a Bayesian Hierarchical Model for Moral Judgments
Final Project for MIT class 6.804: Computational Cognitive Science
pdf -
Minimum Degree of Generating Matrices
Final Project for MIT class 18.821: Seminar in Mathematical Research
pdf -
Methods for Clinical Time Series Analysis in Pediatrics
Final Project for MIT class 6.871: Machine Learning in Healthcare
pdf
Other
In my free time, I enjoy spending my time outdoors hiking/biking/sailing/kayaking in my hometown in the Pacific Northwest and (most recently) Switzerland.