I'm a computer scientist focusing on computer vision and machine learning. My main interests are in geometric deep learning, including both geometry for deep learning (e.g. equivariant neural networks) and deep learning for geometry (e.g. learning-based methods for 3D reconstruction from images).

Currently, I'm a postdoc in the Amsterdam Machine Learning Lab working with Erik J. Bekkers.

I completed my PhD at Chalmers in 2024 under the supervision of Fredrik Kahl, and co-advisors Michael Felsberg and Axel Flinth.

Selected publications

Quick ViTs: Speeding up vision transformers through equivariance
ECCV 2026, arXiv:2505.15441
David Nordström, Johan Edstedt, Fredrik Kahl, Georg Bökman

Identifiable equivariant networks are layerwise equivariant
ICML 2026, arXiv:2601.21645
Vahid Shahverdi, Giovanni Luca Marchetti, Georg Bökman, Kathlén Kohn

Platonic transformers: A solid choice for equivariance
ICML 2026, arXiv:2510.03511
Mohammad Mohaiminul Islam, Rishabh Anand, David R. Wessels, Friso de Kruiff, Thijs P. Kuipers, Rex Ying, Clara I. Sánchez, Sharvaree Vadgama, Georg Bökman, Erik J. Bekkers

Flopping for FLOPs: Leveraging equivariance for computational efficiency
ICML 2025 (Spotlight), arXiv:2502.05169
Georg Bökman, David Nordström, Fredrik Kahl

Steerers: A framework for rotation equivariant keypoint descriptors
CVPR 2024 (Oral), arXiv:2312.02152
Georg Bökman, Johan Edstedt, Michael Felsberg, Fredrik Kahl