Workshop on Machine Learning and Automatic Differentiation in JAX for Scientific Computing
Strasbourg, France
June 8 - 12, 2026
Schedule
9:00
9:30
9:45
10:30
11:00
11:45
12:30
13:00
13:45
14:00
14:30
14:45
15:00
15:30
16:00
16:30
17:00
17:30
9:00 - 9:30
Differentiable simulation with Diffrax: optimal control, inverse problems, and neural ODEs (lecture)
Jonathan Brodrick
9:30 - 10:30
Differentiable simulation with Diffrax: optimal control, inverse problems, and neural ODEs (practical)
Jonathan Brodrick
9:00 - 9:45
Accelerating turbulent simulations of geophysical systems using differentiable programming
Hugo Frezat
9:45 - 10:30
A JAX journey in building TORAX for plasma physics simulation
Sebastian Bodenstein Anushan FernandoCoffee break
11:00 - 12:30
Optimal control of ordinary differential equations (practical)
Killian Lutz Claire Schnoebelen
11:00 - 12:30
PINNs and other neural numerical methods (practical)
Rémi Imbach Victor Michel-Dansac Nilo SchwenckeLunch
Lunch
Lunch
13:45 - 14:00
Welcome
Coffee break
Coffee break
Coffee break
14:30 - 15:00
Building and training end-to-end differentiable systems with the Equinox ecosystem (lecture)
Jonathan Brodrick
15:00 - 16:30
Building and training end-to-end differentiable systems with the Equinox ecosystem (practical)
Jonathan Brodrick
16:00 - 17:30
Differentiable cG/dG methods and numerical linear algebra (practical)
Virgile Bertrand Emmanuel Franck
14:00 - 15:30
The Hybridization of Solvers and Deep Learning with Differentiable Physics (lecture)
Felix Köhler
16:00 - 17:30