AI Winter School 2026 - CFPU/Brown Department of Physics

Jan 6, 2026, 1:00 PM → Jan 9, 2026, 4:30 PM America/New_York

Ian Dell'Antonio (Brown)

Welcome to the 2026 AI Winter School, hosted by the Center for the Fundamental Physics of the Universe, Brown University/Department of Physics

 

The Winter School has concluded. All modules are viewable here on Brown Physics Youtube. For access to module materials, please email cfpu@brown.edu. 

 

This 4-day Winter School is open to interested graduate students, post-doctoral researchers and advanced undergraduates from any institution. Each module and following workshop are intended to stand alone; participants may sign up for as many modules as they choose. 

All modules will take place virtually. There is no registration fee.

Recordings of each module will be available on the Brown Physics YouTube channel. 

 

• View 2026 AI Winter School modules
• View 2025 AI Winter School modules
• View 2024 AI Winter School modules

 

AI Winter School Goals

The goal of this Winter School is to offer examples of the direct applications of machine learning methods to a range of problems inspired by the use of artificial intelligence in physics. We will provide hands-on experiences with machine learning tools to help reduce the barriers to engagement with these essential tools for physics research. 

Hosted by the Brown Center for the Fundamental Physics of the Universe, this year's program will consist of six modules comprised of a lecture session delivered by Brown Physics faculty members and industry experts, followed by practice workshop sessions for participants with direction in the use of machine learning tools.

To learn more about the graduate programs in Physics at Brown, please visit the Physics Department website. 

2026 AI Winter School Program

Roundtable discussion:

Featuring: 

• Brenda Rubenstein, Brown University Professor of Chemistry and Physics and Director of the Brown Data Science Institute
• Michael Littman, Brown University Professor of Computer Science and Associate Provost for Artificial Intelligence
• Phiala Shanahan, MIT Associate Professor of Physics and IAIFI Interim Deputy Director
• Jamie Macbeth, Smith College Associate Professor of Computer Science

Link to Short Description of All Modules

• Module: Introductory Module
• Presenter: Shawn Dubey (Brown University)

 

• Module: Physics‑Informed Neural Networks: Teaching Models the Laws of Nature 
• Presenters: Bryan Ostdiek ((Principal Data Scientist, Microsoft) and Darsh Kodwani 

 

• Module: Applications of Optimal Transport in Collider Physics
• Presenter: Matt LeBlanc (Brown University)

 

• Module: Object Detection for Rare Event Searches
• Presenter: Jeff Schueler (University of New Mexico)

 

• Module: Denoising Electron Densities from Stochastic Simulations of Electronic Structure Using AI
• Presenter: Brenda Rubenstein (Brown University)

 

• Module: Training Simulations to Predict the First Stars and their Effect
• Presenter: Madhurima Choudhury (Birla Institute of Technology & Science Pilani), Jonathan Pober (Brown University)

 

• Module: Reinforcement Learning for Orbital Transfers
• Presenter: James Verbus (Senior Staff Software Engineer, Machine Learning, LinkedIn)

Register for access to all materials

Registration with a Gmail or Google-backed email account is required for full access to all AI Winter School materials including notebooks. Please note that we will use your registration email address to grant access to all materials. Full access to all materials will be available starting January 5. The AI School will make use of Google Colab for all the workbook examples (https://colab.research.google.com/).

If you do not have a Gmail or Google-backed account, you will need to register one.  You can create an account just for this event or perhaps follow the directions at the following link to use an existing email.

Using an Existing Email with Google

NOTE: Module hands-on material will be on a Google shared drive.  Access to this shared drive will be granted to those that have registered through this Indico site with their GMail or Google-backed email account.  Due to the high level of interest this year, registrants will be granted this access via a dedicated Google group to which their registered email will be added.  Access to the shared drive will not be given directly to any registrant's individual account.  If you did not register with a GMail or Google-backed account, please modify or re-register through this site with a GMail or Google-backed account.  If you just registered, please allow a little time for your registered email to be propagated to the Google group.

The Shared Google Drive will continue to be made available beyond the end of the School so there is no need to copy all the contents. Just your working notebook.

The time zone for the school will only be the US East Coast time zone.

Tuesday, January 6

Roundtable Discussion:

https://brown.zoom.us/j/94691206490

Introductory Module

Join here: https://brown.zoom.us/j/94691206490
This module is the introductory module to the virtual AI Winter School. It will consist of a short introductory lecture to AI/ML and then a brief hands-on session that will introduce participants to the basics of using Google Colab, the required platform for this winter school. This module is particularly geared toward those with no knowledge of AI/ML and is recommended as a prerequisite to other modules for those coming in with no background in machine learning.

Convener: Shawn Dubey

Wednesday, January 7

Module 1: Physics‑Informed Neural Networks: Teaching Models the Laws of Nature

Join here: https://brown.zoom.us/j/94691206490
This session introduces Physics‑Informed Neural Networks (PiNNs), a powerful approach that blends data‑driven learning with fundamental physical equations. We’ll explore how incorporating domain knowledge into the training process improves accuracy, reduces data requirements, and increases interpretability. Through intuitive examples and hands‑on exercises, participants will learn how PiNNs solve differential equations, model real‑world systems, and outperform traditional ML approaches in low‑data regimes.

Convener: Bryan Ostdiek (Microsoft)

Module 2: Applications of Optimal Transport in Collider Physics

Join here: https://brown.zoom.us/j/94691206490
Optimal Transport (OT) has emerged as a powerful framework in high-energy physics that offers a geometrical language for data analysis. This module explores applications of OT to collider physics, focused on the "Energy-Mover's Distance." We will explore the metric space of particle physics data, and learn how many aspects of collider data analysis like jet clustering and event shape definitions can be recast in this common language. Through this module, participants will gain intuition for this approach and practical experience with common libraries like Python Optimal Transport that is necessary to leverage OT in large-scale data analysis.

Convener: Matt LeBlanc (Brown)

Thursday, January 8

Module 3: Training Simulations to Predict the First Stars and their Effect

Join here: https://brown.zoom.us/j/94691206490
We are interested in regression tasks in analyzing 21 cm cosmology data. Most semi-numerical algorithms use a unique (and not always physical) parameterization of the 21 cm signal, but we can regress on well-defined phenomenological parameters. In this module, we’ll look at estimating the neutral fraction of the intergalactic medium from 21 cm signal maps.

Conveners: Jonathan Pober (Brown), Madhurima Choudhury

Module 4: Object Detection for Rare Event Searches

Join here: https://brown.zoom.us/j/94691206490
Object detection is a cross-disciplinary computer vision technique where a model is trained to simultaneously classify and localize objects of interest in image data. This module introduces the notion of object detection and how the MIGDAL experiment uses it to enable real data training in its search for the Migdal effect. While we use simulation in this module, participants will be given hands-on examples demonstrating the usage of object detection to reconstruct rare events as well as a brief introduction to labeling software.

Convener: Jeffrey Schueler (University of New Mexico)

Friday, January 9

Module 5: Denoising Electron Densities from Stochastic Simulations of Electronic Structure Using AI

Join here: https://brown.zoom.us/j/94691206490
One of the most computationally-efficient approaches for modeling solids that feature strong electron correlation is by using Quantum Monte Carlo techniques. These techniques are among the most accurate available, but importantly, scale gracefully with system size because of their use of random sampling. Nonetheless, this random sampling can make the measurement of observables such as the electron charge or spin density noisy, requiring long averaging times. Here, we show how ML methods for the denoising of images can be used to denoise stochastic simulations of the electron density.

Convener: Brenda Rubenstein (Brown)

Module 6:: Module 6: Reinforcement Learning for Orbital Transfers

Join here: https://brown.zoom.us/j/94691206490
Using a simple 2D orbital transfer as a testbed, we’ll compare an analytic Hohmann solution to a controller learned from trial-and-error in simulation. You’ll build the RL setup from scratch (state/action/reward), train a PPO policy, and develop intuition by inspecting the learned trajectories, trade-offs, and characteristic failure cases.

Convener: James Verbus