MatrixNet: Learning over symmetry groups using learned group representations
Not on the map. Why?
Abstract (arXiv)
Group theory has been used in machine learning to provide a theoretically grounded approach for incorporating known symmetry transformations in tasks from robotics to protein modeling. In these applications, equivariant neural networks use known symmetry groups with predefined representations to learn over geometric input data. We propose MatrixNet, a neural network architecture that learns matrix representations of group element inputs instead of using predefined representations. MatrixNet achieves higher sample efficiency and generalization over several standard baselines in prediction tasks over the several finite groups and the Artin braid group. We also show that MatrixNet respects group relations allowing generalization to group elements of greater word length than in the training set.
Key facts. Cited by 2 papers in the map (data through 7 Oct 2026). Builds on 29 of its 48 references that are in the map. First posted on arXiv 16 Jan 2025.
Builds on
29 of 48 references.
References · 29
- 2017Attention Is All You Needcited by 59,932
- 2020An Image is Worth 16x16 Words: Transformers for Image Recognition at Scalecited by 29,915
- 2016Semi-Supervised Classification with Graph Convolutional Networkscited by 11,282
- 2018On the Continuity of Rotation Representations in Neural Networkscited by 1,186
- 2016Group Equivariant Convolutional Networkscited by 1,154
- 2013Spectral Networks and Locally Connected Networks on Graphscited by 1,097
- 2022FourCastNet: A Global Data-driven High-resolution Weather Model using Adaptive Fourier Neural Operatorscited by 748
- 2018Tensor field networks: Rotation- and translation-equivariant neural networks for 3D point cloudscited by 735
- 2017Physics Informed Deep Learning (Part I): Data-driven Solutions of Nonlinear Partial Differential Equationscited by 362
- 2018On the Generalization of Equivariance and Convolution in Neural Networks to the Action of Compact Groupscited by 296
- 2022Learning Local Equivariant Representations for Large-Scale Atomistic Dynamicscited by 254
- 2019Contrastive Learning of Structured World Modelscited by 206
17 more
- 2020Sampling using SU(N) gauge equivariant flowscited by 83
- 2021Automatic Symmetry Discovery with Lie Algebra Convolutional Networkcited by 82
- 2022SO(2)-Equivariant Reinforcement Learningcited by 79
- 2021Deep Regression on Manifolds: A 3D Rotation Case Studycited by 67
- 2018Explorations in Homeomorphic Variational Auto-Encodingcited by 65
- 2021Implicit-PDF: Non-Parametric Representation of Probability Distributions on the Rotation Manifoldcited by 65
- 2022Learning Symmetric Embeddings for Equivariant World Modelscited by 54
- 2022Equivariant Descriptor Fields: SE(3)-Equivariant Energy-Based Models for End-to-End Visual Robotic Manipulation Learningcited by 54
- 2023Self-supervised learning of Split Invariant Equivariant representationscited by 41
- 2023Equivariant Adaptation of Large Pretrained Modelscited by 40
- 2024Learning with 3D rotations, a hitchhiker's guide to SO(3)cited by 39
- 2023Latent Space Symmetry Discoverycited by 29
- 2024Mechanistic Neural Networks for Scientific Machine Learningcited by 16
- 2020ForecastNet: A Time-Variant Deep Feed-Forward Neural Network Architecture for Multi-Step-Ahead Time-Series Forecastingcited by 7
- 2022Differentiable Mathematical Programming for Object-Centric Representation Learningcited by 6
- 2023Learning Lie Group Symmetry Transformations with Neural Networkscited by 6
- 2020Algebraically-Informed Deep Networks (AIDN): A Deep Learning Approach to Represent Algebraic Structurescited by 2
Built on by
Citing papers per year: 2025: 2.
2 papers.