Clebsch-Gordan Transformer: Fast and Global Equivariant Attention

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Howell, Owen Lewis, Zhao, Linfeng, Zhu, Xupeng, Qian, Yaoyao, Huang, Haojie, Sun, Lingfeng, Thomason, Wil, Platt, Robert, Walters, Robin
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916975037382656
author Howell, Owen Lewis
Zhao, Linfeng
Zhu, Xupeng
Qian, Yaoyao
Huang, Haojie
Sun, Lingfeng
Thomason, Wil
Platt, Robert
Walters, Robin
author_facet Howell, Owen Lewis
Zhao, Linfeng
Zhu, Xupeng
Qian, Yaoyao
Huang, Haojie
Sun, Lingfeng
Thomason, Wil
Platt, Robert
Walters, Robin
contents The global attention mechanism is one of the keys to the success of transformer architecture, but it incurs quadratic computational costs in relation to the number of tokens. On the other hand, equivariant models, which leverage the underlying geometric structures of problem instance, often achieve superior accuracy in physical, biochemical, computer vision, and robotic tasks, at the cost of additional compute requirements. As a result, existing equivariant transformers only support low-order equivariant features and local context windows, limiting their expressiveness and performance. This work proposes Clebsch-Gordan Transformer, achieving efficient global attention by a novel Clebsch-Gordon Convolution on $\SO(3)$ irreducible representations. Our method enables equivariant modeling of features at all orders while achieving ${O}(N \log N)$ input token complexity. Additionally, the proposed method scales well with high-order irreducible features, by exploiting the sparsity of the Clebsch-Gordon matrix. Lastly, we also incorporate optional token permutation equivariance through either weight sharing or data augmentation. We benchmark our method on a diverse set of benchmarks including n-body simulation, QM9, ModelNet point cloud classification and a robotic grasping dataset, showing clear gains over existing equivariant transformers in GPU memory size, speed, and accuracy.
format Preprint
id arxiv_https___arxiv_org_abs_2509_24093
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Clebsch-Gordan Transformer: Fast and Global Equivariant Attention
Howell, Owen Lewis
Zhao, Linfeng
Zhu, Xupeng
Qian, Yaoyao
Huang, Haojie
Sun, Lingfeng
Thomason, Wil
Platt, Robert
Walters, Robin
Machine Learning
Computer Vision and Pattern Recognition
Robotics
The global attention mechanism is one of the keys to the success of transformer architecture, but it incurs quadratic computational costs in relation to the number of tokens. On the other hand, equivariant models, which leverage the underlying geometric structures of problem instance, often achieve superior accuracy in physical, biochemical, computer vision, and robotic tasks, at the cost of additional compute requirements. As a result, existing equivariant transformers only support low-order equivariant features and local context windows, limiting their expressiveness and performance. This work proposes Clebsch-Gordan Transformer, achieving efficient global attention by a novel Clebsch-Gordon Convolution on $\SO(3)$ irreducible representations. Our method enables equivariant modeling of features at all orders while achieving ${O}(N \log N)$ input token complexity. Additionally, the proposed method scales well with high-order irreducible features, by exploiting the sparsity of the Clebsch-Gordon matrix. Lastly, we also incorporate optional token permutation equivariance through either weight sharing or data augmentation. We benchmark our method on a diverse set of benchmarks including n-body simulation, QM9, ModelNet point cloud classification and a robotic grasping dataset, showing clear gains over existing equivariant transformers in GPU memory size, speed, and accuracy.
title Clebsch-Gordan Transformer: Fast and Global Equivariant Attention
topic Machine Learning
Computer Vision and Pattern Recognition
Robotics
url https://arxiv.org/abs/2509.24093