Workshop on Machine Learning and Automatic Differentiation in JAX for Scientific Computing

Strasbourg, France
June 8 - 12, 2026

Provisional schedule

9:00

9:30

10:30

11:00

11:45

12:30

13:00

13:45

14:00

14:30

15:30

16:00

16:45

17:30

9:00 - 10:30 Pierre Glaser

Introduction to the JAX framework

9:00 - 9:30 Patrick Kidger

The diffrax library: ODEs and neural ODEs

9:30 - 10:30

Practical session:
The diffrax library

instructor: Patrick Kidger
9:00 - 10:30 Victor Michel-Dansac

PINNs and other neural numerical methods

9:00 - 9:45 Hugo Frezat

Climate modeling

9:45 - 10:30 Claire Schnoebelen

Neural flow for Hamiltonian problems

Coffee break

11:00 - 13:00

Practical session:
introduction to the JAX framework

instructor: Pierre Glaser
11:00 - 12:30

Practical session:
optimal control of ODEs

instructors: Killian Lutz &
Claire Schnoebelen
11:00 - 12:30

Practical session:
PINNs and other neural numerical methods

instructors: Rémi Imbach & Victor Michel-Dansac
11:00 - 11:45 Deniz Bezgin &
Aaron Buhendwa

The JAX-Fluids library

11:45 - 12:30 Nora Loose

Climate modeling

Lunch

Lunch

Lunch

13:45 - 14:00

Welcome

14:00 - 15:30 Samuel Vaiter

Introduction to automatic differentiation

Coffee break

16:00 - 17:30

Practical session:
automatic differentiation

instructor: Samuel Vaiter
14:30 - 15:00

Introduction to neural networks

15:00 - 15:30 Patrick Kidger

Neural networks in JAX: the equinox library

16:00 - 17:30

Practical session:
the equinox library

instructor: Patrick Kidger
14:00 - 15:30

Practical session:
differentiable numerical linear algebra

instructor: Emmanuel Franck
16:00 - 16:45 Roxana Sublet

Particle simulations

16:45 - 17:30 Jonathan Citrin

The TORAX code for plasma physics

14:00 - 15:30 Felix Koehler

Hybridization of classical solvers and neural methods

16:00 - 17:30

Practical session:
hybrid methods

instructor: Felix Koehler