Jonathan Gornet

About

I am an AI researcher and Ph.D. candidate in Systems Science and Mathematics at Washington University in St. Louis specializing in decision-making under uncertainty for dynamic environments, with applications in automated neural network training for reinforcement learning and generative AI.

My experience progresses from my current graduate research developing novel machine learning models and decision-making algorithms, to previous roles at In-Q-Tel Lab41 (anomaly detection in deep learning systems), HHMI Janelia (optimizing large-scale simulations), MITRE (NLP for trajectory prediction), and NYU (optimization methods for networked systems).

Published in conferences including L4DC and IEEE Control and Decision, I combine strong technical skills in Python (PyTorch, JAX), automated machine learning, and statistical modeling with a strong mathematics background from Washington University in Saint Louis and NYU. My work bridges theoretical research with practical applications to create innovative AI solutions for complex challenges.

Research Interests:

Data-Driven OptimizationReinforcement LearningAutomated Machine Learning
Jonathan Gornet

Education

  1. Washington University in Saint Louis

    Ph.D. in Systems Science & Mathematics

  2. New York University

    Bachelors of Arts in Mathematics

  3. Loyola University Chicago

    Pre-med coursework

Experience

  1. Washington University in Saint Louis

    Ph.D. Candidate

    Researching optimal decision-making in unknown and stochastically changing environments, with applications in automated neural network training.

  2. IQT Lab41

    Data Scientist Intern

    Developed novel methods for anomaly detection in deep learning systems' behavior in deployed environments.

  3. HHMI Janelia Research Campus

    Undergraduate Scholar

    Improved efficiency of large-scale simulations modeling a fly brain's visual system as a dynamical system.

  4. New York University

    Undergraduate Researcher

    Implemented optimization methods for controlling behavior of networked systems, including modeling biological brain activity during sleep.

  5. MITRE

    Aviation Systems Analyst Intern

    Applied natural language processing to predict flight trajectories using flight tower commands.

Publications

  1. Jonathan Gornet and Bruno Sinopoli. "Restless Bandit Problem with Rewards Generated by a Linear Gaussian Dynamical System". Learning for Dynamics and Control 2024.

  2. Jonathan Gornet, Mehdi Hosseinzadeh, and Bruno Sinopoli. "Stochastic multi-armed bandits with non-stationary rewards generated by a linear dynamical system". IEEE Conference on Decision and Control 2022.

  3. Jonathan Gornet, Mehdi Hosseinzadeh, and Bruno Sinopoli. "An Adaptive Method for Non-Stationary Stochastic Multi-armed Bandits with Rewards Generated by a Linear Dynamical System". ArXiv 2024.

  4. Jie Wang, Jonathan Gornet, Alex Orange, Leigh Stoller, Gary Wong, Jacobus Van Der Merwe, Sneha Kumar Kasera, Neal Patwari. "Two Measure is Two Know: Calibration-free Full Duplex Monitoring for Software Radio Platforms". IEEE DySPAN 2024.

  5. Jonathan Gornet and Louis Scheffer. "Simulating extracted connectomes". BioRxiv 2018.

  6. Daniel Levenstein, Jonathan Gornet, Roman Huszár, Gabrielle Girardeau, Andres Grosmark, Adrien Peyrache, Yuta Senzai, Brendon O Watson, Kenji Mizuseki, John Rinzel, György Buzsáki. "Distinct ground state and activated state modes of firing in forebrain neurons". BioRxiv 2021.