Skip to the content.
Submit to our competition track! Deadline: April 22, 2026
Submit to our blogpost track! Deadline: April 15, 2026

We are pleased to announce this second edition of GRaM, as an ICLR 2026 workshop. This year, we will have a focus on scale and simplicity. We open three tracks: paper tracks, blogposts track and a new competition track. You can find us on the official ICLR webpage.

Deadlines

Call for Blogposts (Submit)

We invite submissions to our blogpost track. We welcome blog posts in the ICLR Distill format that communicate ideas clearly and accessibly. See the full guidelines and submit on the blogpost website, and explore blog posts from GRaM 2024 for examples. Deadline: April 15, 2026 (AoE).

Call for Competition (Submit)

We are hosting a benchmark challenge on transient airflow prediction over 3D geometries inspired by Formula 1 front wings, with simulation data provided by BeyondMath. Given a geometry and airflow at previous time points, predict the airflow at future time points. The winner receives the MCML Award (500 € prize). All valid submissions will be published in the workshop proceedings with co-authorship. Submissions are made via pull requests — see the competition website for details. Deadline: April 22, 2026 (AoE).

Accepted Papers (OpenReview)

The call for papers is closed. You can see all the accepted papers on the openreview page ICLR 2026 Workshop GRaM. Congratulations to all the authors!

Motivation

Many real-world datasets have geometric structure, but most ML methods ignore such structure, and treat all inputs as plain vectors. GRaM is a workshop about grounding models in geometry, using ideas from group equivariance to non-Euclidean metrics, to build better, more interpretable representations and generative models.

An approach is geometrically grounded if it respects the geometric structure of the problem domain and supports geometric reasoning.

For this second edition, we aim to explore the relevance of geometric methods, particularly in the context of large models, focusing on the theme of scale and simplicity.

Topics

We solicit submissions that present theoretical research, methodologies, applications, insightful analysis, and even open problems, within the following topics (list not exhaustive):

Organizers

Alison Pouplin Alison Pouplin
Bayer
Sharvaree Vadgama Sharvaree Vadgama
UC San Diego
Erik Bekkers Erik Bekkers
Universiteit van Amsterdam
Sékou-Oumar Kaba Sékou-Oumar Kaba
McGill University and Mila
Hannah Lawrence Hannah Lawrence
MIT
Manuel Lecha Manuel Lecha
IIT and Oxford University
Elizabeth (Libby) Baker Elizabeth (Libby) Baker
DTU Denmark
Julian Suk Julian Suk
TU Munich
Robin Walters Robin Walters
Northeastern University
Jakub Tomczak Jakub Tomczak
Chan Zuckerberg Initiative
Stefanie Jegelka Stefanie Jegelka
TU Munich

Workshop Sponsor

New Theory

Competition Sponsors

BeyondMath MCML

Acknowledgements

We are grateful to New Theory for their generous financial support of the workshop. We also thank BeyondMath for creating and curating the incredible dataset powering our competition, with a special thanks to Gavin Seegoolam (BeyondMath) and Julian Suk (Competition Area Chair) for making it happen. Finally, we thank MCML for providing the competition winners with an award prize.

Area Chairs

We would like to thank our Area Chairs for their invaluable help with the review process: