"False"
Skip to content
printicon
Main menu hidden.

GeUmetric Deep Learning Workshop 2026

Program

Day 1, Monday 17/8

13:00-13:30 Welcome

13:30-14:00 Hampus Linander (Chalmers University of Technology)
Soft geometric inductive bias

14:00-14:30 Jakob Galley (Umeå University)
Conservation Laws from Data Symmetry in Neural Networks

14.30-15:00 Isabel Dahlgren (ETH Zürich)
Deep, but not difficult: identifiability and critical points for residual networks

15:00-15:30 Coffee break

15:30-16:00 Viktor Vigren Näslund (Umeå University)
PolyNODE: Variable-dimension Neural ODEs on M-polyfolds

16:00-16:30 Longde Huang (Chalmers University of Technology)
Boosting Data Augmentation with Stochastic Weight Averaging

16:30-17:00 Thomas Grill (KTH Royal Institute of Technology)
The Boundary of Neuromanifolds of shallow MLPs

Day 2, Tuesday 18/8

09:00-10:00 Juan Viguera (Astra Zeneca)
Generative molecular dynamics and Implicit Transfer Operators

10:00-10:30 Coffee break

10:30-11:00 Emma Andersdotter (Umeå University)
Steerable Neural ODEs on Homogeneous Spaces

11:00-11.30 Jiri Minarcik (Carnegie Mellon/Imperial College)
Untangling Surfaces via Shape and Mesh Repulsion

11:30-13:00 Lunch

13:00-14:00 Vishal Ngairangbam (Karlsruhe Institute of Technology)
Theory-priors for Deep Learning in Collider Phenomenology

14:00-14:30 Johanna Marie Gegenfurtner (Technical University of Denmark)
The Symmetries of Three-Layer ReLU Networks

14:30-15:00 Li Ju (Uppsala University)
Geometry-Aware Uncertainty Quantification for Frozen Vision–Language Models

15:00-15:30 Coffee break

15:30-16:00 Philipp Misof (Chalmers University of Technology)
Finite-Width Neural Tangent Kernels from Feynman Diagrams

16:00-16:30 Johan Henriksson (Umeå University)
Gene function prediction - Challenges and opportunities

16:30-17:00 Johannes Maeß  (TU Berlin)
Implicit Machine Learning Force Fields: Fast Dynamics for Molecular Geometries

Day 3, Wednesday 19/8

09:00-10:00 Andreas Ipp (TU Wien)
Lattice Gauge-Equivariant Neural Networks: From Fixed-Point Actions to Diffusion Models

10:00-10:30 Coffee break

10:30-11:00 Miaowen Dong (Chalmers University of Technology)
Equivariance and Augmentation for Bayesian Neural Networks

11:00-11.30 Aron Persson (Umeå University)
Stochastic currents as diffusion laws

11:30-12:00 Oguzhan Recep Akkol (Aix-Marseille University)
Detecting ADE Singularities via Realization Fibers: Toward Explainable Polynomial Neural Networks

 

Book of Abstracts GeUmetric 2026

GeUmetric Deep Learning Workshop 2026

Latest update: 2026-08-21