Proceedings Track Paper
Full papers up to 8 pages (excluding references and appendices); accepted papers are published in PMLR.
Machine Learning/CORE Unranked
ICLR 2026
ICLR 2026 Workshop on Geometry-grounded Representation Learning and Generative Modeling
The GRaM Workshop @ ICLR 2026 is a venue for research on Geometry-grounded Representation Learning and Generative Modeling, with a focus on scale and simplicity. It brings together researchers to explore how geometric principles intersect with modern machine learning, particularly in areas like equivariance, topological deep learning, and scalable architectures.
Key deadlines
Verified September 9, 2026Full paper
February 6, 2026
AoE
Workshop timeline
Submission and decisions
Full paperKey deadline
February 6, 2026 · AoE
Paper fit
Contribution paths
A strong submission should clearly identify its contribution and evaluate it appropriately.
Extended Abstract
Short papers up to 4 pages (excluding references and appendices); non-archival, available on OpenReview but not in formal proceedings.
Blogpost
Non-archival blog posts up to 2000 words in markdown format, with tutorial value, opinion pieces, open problems, or reproducibility results.
Tutorial (Colab Notebook)
Collaborative notebook submissions with checkpoints for reproducibility, focused on tutorial value for GRaM-relevant topics.
Topological Deep Learning Challenge
Participation in the TAG challenge to design lifting mappings between data structures and topological domains (e.g., point-clouds, hypergraphs, simplicial complexes).
Research areas in scope
01
Generative Modeling
Generative modeling
02
Representation Learning
Representation learning
03
Molecular / Scientific / 3D
Molecular / Scientific / 3D
04
LLMs / Foundation Models
LLMs / Foundation models
05
Optimization / Training
Optimization / Training
06
Theory / Expressivity
Theory / Expressivity
07
Scaling / Efficiency
Scaling / Efficiency
08
Inductive Bias / Architecture
Inductive / Architecture
09
Graph Theory
Graph theory
10
Group Theory / Equivariance
Group theory / Equivariance
11
Differential Geometry
Differential geometry
12
Topology / TDA
Topology / TDA
13
Hyperbolic / Non-Euclidean Spaces
Hyperbolic / Non-Euclidean spaces
14
Optimal Transport
Optimal transport
15
Linear Algebra / Spectral Methods
Linear algebra / Spectral methods
16
Information Theory / Dimensionality
Information theory / Dimensionality
17
Empirical / Structured Data
Empirical / Struct. data
Policies worth checking twice
- Papers in the Proceedings Track must include at least 30% new, unpublished/unsubmitted material if they overlap with content submitted or published elsewhere.
- To publish a GRaM paper in another venue later, authors must add at least 30% new material.
- Extended abstracts and blogposts are non-archival and do not appear in formal proceedings.
- Blogpost submissions must be in provided markdown format.
- Tutorial submissions must be Colab notebooks with checkpoints for quick reproducibility.
Official sources
Compiled from the official call for papers. The organizers’ pages remain authoritative.
Last verified September 9, 2026