ICLR 2026PastGenerative models
ICLR 2026 Workshop on Geometry-grounded Representation Learning and Generative Modeling
ICLR 2026 Workshop GRaM
- Submission deadline
- Feb 6, 2026, 12:00 UTCOpenReview-synced 2026-02-06 12:00 UTC (as of 2026-06-23) — extensions on OpenReview are applied automatically; verify on the website.
- Submission portal
- OpenReview
- Notes
- Topics were auto-suggested and may be imprecise — edits welcome.
Accepted papers (83)
Fetched from OpenReview (v2) on 2026-06-10.
A Geometric Perspective on the Difficulties of Learning GNN-based SAT Solvers
A Graph-Theoretical View of Space Folding via the Motzkin–Straus Framework
A universal compression theory for lottery ticket hypothesis and neural scaling laws
Adaptive Quasimetric Mapping : Principled Topological Abstraction for Robust Offline Goal-Conditioned Navigation
Adaptive Symmetry Discovery for Dynamical System Identification
ADAPTIVEMIXGNN: Local Adaptive Inductive Bias for Heterophilic Node Classification
Algebraic priors for approximately equivariant networks
Autoregressive Frontier Expansion: Growing Trees with Graph Machine Learning
Balancing Symmetry and Efficiency in Graph Flow Matching
Beyond Co-occurence: A Study of Early-stage Semantic Geometry in Next-Token Prediction
Beyond Linearity in Attention Projections: The Case for Nonlinear Queries
Can Graph Foundation Models Generalize Over Architecture?
Categorical Trace Loop Networks for Gauge-Randomized Holonomy Regression
CLERF: Contrastive LEaRning for Full-Range Head Pose Estimation
Conformal Coordinate Frames for Disentanglement
Data-Adaptive Relaxed Equivariant Networks for Symmetry Breaking
DiScoFormer: Plug-In Density and Score Estimation with Transformers
DO CORESETS, PRUNING, AND QUANTIZATION PRESERVE NEURAL NETWORK REPRESENTATIONS?
E$(n)$-Equivariant Spherical Decision Surfaces
Effective Resistance Rewiring: A Simple Topological Correction for Over-Squashing
Embedding Compression via Spherical Coordinates
Eq-WaLa: Equivariant Augmentation and Regularization for Wavelet Latent Flow Matching
Flow curvature explains failed SDE drift estimation under sparse sampling
Fréchet Regression on the Bures-Wasserstein Manifold
From Leads to Latents: Attention-Driven Masked Autoencoder for ECG Time Series
Generalized Reduction to the Isotropy for Flexible Equivariant Neural Fields
Geometric Inductive Biases for Diffusion-Based Graph Generation
Geometry-Driven Diverse and Transferable Visual Attacks on Multimodal LLMs
Geometry-Grounded Flow Matching on Compact Manifolds
GSVD for Geometry-Grounded Dataset Comparison: An Alignment Angle Is All You Need
Hyperbolic Curvature as an Inductive Bias for Latent Space Flow Matching
Hyperbolic Geometry of Reasoning: Probing LLM Hidden States
Improving LLM Predictions via Inter-Layer Structural Encoders
InertialAR: Autoregressive 3D Molecule Generation with Inertial Frames
INTRINSIC DIMENSION DYNAMICS IN ACTIVE LEARNING: A GEOMETRIC DIAGNOSTIC OF ACQUISITION BEHAVIOR
k-Maximum Inner Product Attention for Graph Transformers and the Expressive Power of GraphGPS
Laplacian Flows for Policy Learning from Experience
Latent Equivariant Operators for Robust Object Recognition: Promises and Challenges
Learning Compact Representations via Intrinsic Dimension Regularization
Learning in Transformers under Spectral Constraints
Lift me up: the impact of liftings on hypergraph neural networks
LIGHT CONES FOR VISION: SIMPLE CAUSAL PRIORS FOR VISUAL HIERARCHY
Manifold Generalization Provably Proceeds Memorization in Diffusion Models
Metric multi-dimensional scaling for longitudinal data embeddings in pharmacometrics
mHC-lite: You Don't Need 20 Sinkhorn-Knopp Iterations
Mix Early, Forget Less: Data Mixing During Pretraining Builds Resistance to Forgetting
Mutual Information and Task-Relevant Latent Dimensionality
Neurodiversity Meets Colors: Does Position Awareness Destroy Generalization in Brain Graph Learning?
On Closed-Form Couplings
On the Expressive Power of Mixed-Curvature Representations in Product Manifolds
On the Fisher Geometry of Diffusion Models' Latent Space
On the Geometry of Analogical Reasoning in Latent Space
On the necessity of learnable sheaf laplacians
Operator-Consistent Graph Neural Networks for Learning Diffusion Dynamics on Irregular Meshes
Orthogonal Self-Attention
Pawsterior: Variational Flow Matching for Structured Simulation-Based Inference
Physics-Aligned Decoding (PAD) for Discrete Protein Structure Representations
Platonic Transformers: A Solid Choice for Equivariance
Poisson-Induced Potentials for Contractive representations
ProCLIP: Product Space Multimodal Contrastive Alignment
Random but Right: A Geometric Explanation for Efficient LLM Training
RECYCLE NET: CYCLE-AWARE, FEATURE-FREE GNN FOR COMMUNITY DETECTION
Riemannian Metric Matching for Scalable Geometric Modelling of Distributions
Rigid Invariant Sliced Wasserstein via Independent Embeddings
Scalable Message Passing Neural Networks: No Need for Attention in Large Graph Representation Learning
Scale Continuity in Graph Learning: Going beyond spectral methods
Semantic-Anchored, Class Variance-Optimized Clustering for Robust Semi-Supervised Few-Shot Learning
Sharpness-Aware Pretraining Mitigates Catastrophic Forgetting
Solvaformer: Minimizing Geometric Redundancy for Scalable Solubility Prediction
Sparse Concept Anchoring for Interpretable and Controllable Neural Representations
Spatio-Spectral Sequence Processing
Symmetry, Gauss-Newton, and Whitening in Neural Network Optimization
Tensor-SAE: Structured Sparse Autoencoders for Interpretable and Efficient Image Representations
The Affine Divergence: Aligning Activation Updates Beyond Normalisation
The Geometrical and Topological Signature of Transformers
The Geometry of Spectral Gradient Descent: Layerwise Criteria for SignSGD vs SpecSGD
Topological Invariance and Breakdown in Learning
TopoPointPWC: Manifold Topology-Aware Point Cloud Registration via Persistent Homology
Towards a Geometric Theory of Fairness: Detecting Mode Collapse on the Grassmannian Manifold
Towards Scalable Persistence-Based Topological Optimization
Towards Text-Line Segmentation of Historical Documents Using Graph Neural Networks
TPR-Attention for Combinatorial Generalization
Weak-SIGReg: Covariance Regularization for Stable Deep Learning