Joey Rudoler
I’m a PhD student at The Wharton School, where I study Statistics and Data Science. My interests span deep learning theory, neuroscience, and the statistics of both.
I was drawn to statistics and machine learning by a desire to find my place in a world overwhelmed with data. There are lots of problems I’d like to tackle (probably too many), and you can read about some of my past work and current interests on the research page. I love thinking about the world both experientially and mathematically, and I honestly believe we need both perspectives to have any idea what’s going on around us.
Background
Before starting my PhD, I worked as a software engineer facilitating computational cognitive neuroscience research at Penn’s Computational Memory Lab. While working for the lab I finished a part-time engineering master’s in Data Science, which included a thesis supervised by Konrad Kording. I attended Penn for undergrad as well, majoring in Physics & Astronomy with a Concentration in Business and Technology, along with minors in English and Music.
Research
My research has wandered from machine learning applications in neuroscience toward the theory and methodology of deep learning. Across both areas, I'm interested in how learning systems pick up useful patterns, and when those patterns carry over to new people, data, or tasks.
Research notes & figures · All publications
Selected work
Weight Space Gradient Descent Induces Approximate Activity Descent, Making Learning Rules Testable
NeurIPS 2026 Workshop on Symmetry and Geometry in Neural Representations (NeurReps), Extended Abstract Track
Decoding EEG for Optimizing Naturalistic Memory
Journal of Neuroscience Methods