Josh Vendrow

EECS PhD at MIT

About Me

I am a second-year PhD student at MIT EECS, where I am advised by Aleksander Mądry. I completed my BS double majoring in Computer Science and Math at UCLA, where I was advised by Deanna Needell and mentored by Jamie Haddock and Hanbaek Lyu.

My research focuses on principled ways to make machine learning models safe and robust when deployed in the real world. During my undergrad, I worked on on applications of matrix and tensor factorization to topic modeling, computer vision, and network science.

Papers

* denotes equal contribution

Ask Your Distribution Shift if Pre-Training is Right for You
Benjamin Cohen-Wang, Josh Vendrow, Aleksander Madry
Distribution Shifts (DistShift) Workshop, NeurIPS, 2023
[arXiv] [Blog 1] [Blog 2] [Code]

The Journey, Not the Destination: How Data Guides Diffusion Models
Kristian Georgiev*, Josh Vendrow*, Hadi Salman, Sung Min Park, Aleksander Madry
Challenges of Deploying Generative AI Workshop, ICML, 2023
[arXiv] [Blog] [Code]

Dataset Interfaces: Diagnosing Model Failures Using Controllable Counterfactual Generation
Josh Vendrow*, Saachi Jain*, Logan Engstrom, Aleksander Madry
Data-centric Machine Learning Research (DMLR) Workshop, ICML, 2023
[arXiv] [Blog] [Code]

Learning low-rank latent mesoscale structures in networks
Hanbaek Lyu, Yacoub Kureh, Josh Vendrow, M. A. Porter
Nature Communications, 2023 [Journal] [arXiv] [Code]

A Generalized Hierarchical Nonnegative Tensor Decomposition
Josh Vendrow, Jamie Haddock, and Deanna Needell
IEEE Int. Conf. on Acoustics, Speech, Signal Processing (ICASSP), 2022
[Proceedings] [arXiv] [Code]

Learning to predict synchronization of coupled oscillators on heterogeneous graphs
Hardeep Bassi, Richard P. Yim, Rohith Kodukula, Josh Vendrow, Cherlin Zhu, Hanbaek Lyu
Scientific Reports, 2022 [Journal] [arXiv] [Code]

Realistic Face Reconstruction from Deep Embeddings
Edward Vendrow, Josh Vendrow
Privacy in Machine Learning Workshop, NeurIPS, 2021 [Code]

On a Guided Nonnegative Matrix Factorization
Josh Vendrow, Jamie Haddock, Elizaveta Rebrova, Deanna Needell
IEEE Int. Conf. on Acoustics, Speech, Signal Processing (ICASSP), 2021
[Proceedings] [arXiv] [Code] [Poster] [Slides]

Neural Nonnegative CP Decomposition for Hierarchical Tensor Analysis.
Josh Vendrow, Jamie Haddock, Deanna Needell.
Asilomar Conf. on Signals, Systems and Computers, 2021
[Code] [Poster] [Slides]

On the Relation of Gene Essentiality to Intron Structure: A Computational and Deep Learning Approach
Ethan Schonfeld, Edward Vendrow, Josh Vendrow, Elan Schonfeld
Life Science Alliance, 2021 [Journal] [bioRxiv] [Code]

Feature Selection from Lyme Disease Patient Survey Data Using Machine Learning
Josh Vendrow, Jamie Haddock, Deanna Needell, Lorraine Johnson
Algorithms, 2020 [Journal] [arXiv] [Code]

Antibiotic Treatment Response In Persistent Lyme Disease: Why Do Some Patients Improve While Others Do Not?
Lorraine Johnson, Mira Shapiro, Raphael Stricker, Josh Vendrow, Jamie Haddock, Deanna Needell
Healthcare, 2020 [Journal]

Software Development

fnnls
An implementation of the Fast Nonnegative Least Squares algorithm.
[PyPi] [Code]

Network Dictionary Learning
Learning from and reconstructing networks using MCMC motif sampling and Nonnegative Matrix Factorization.
[PyPi] [Code]

NNetwork
A custom graph/network/multi-weighted network class optimized for scalable sampling and searching algorithms.
[PyPi] [Code]

Twin
I also have a twin brother! Check out Edward Vendrow.