Symmetry-Preserving Pruning of Group Equivariant Convolutional Networks via Representation Theory
Symmetry · 06.06.2026 · 18 (6) · 983
DOI: 10.3390/sym18060983 ↗Kurz gefasst
Fachartikel in der Zeitschrift Symmetry, online erschienen am 06.06.2026. Verfasst von Mohammed Alnemari und 1 weiteren Autor:innen. Die Arbeit nennt 41 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: King Faisal University.
Abstract (Originalsprache)
Group equivariant convolutional neural networks (G-CNNs) achieve superior sample efficiency by encoding symmetry into network architecture, yet their computational overhead (up to 3.78× slower inference and 4.63× more multiply–accumulate operations) hinders deployment on resource-constrained edge devices. Existing pruning methods cannot be applied directly: arbitrarily removing weights breaks the group representation structure and degrades equivariance. We characterize the complete design space of equivariance-preserving compression, proving that exactly two axes leave a convolutional layer equivariant: irrep-bundle pruning, which reduces irreducible-representation multiplicities, and orbit-wise pruning, which removes complete spatial orbits from kernel supports; via Schur’s lemma, no third structure-preserving axis exists. This completeness result, rather than the use of representation theory itself, is our central contribution. We turn it into practice through direct sub-filter extraction, which yields real convolutional parameter reduction (up to 83%) and 1.4–2.9× measured inference speedup, unlike masking, which gives no real speedup. Across three datasets (MNIST, CIFAR-10, EuroSAT) and three symmetry groups (C4, D4, SO(2)), compression is nearly lossless on strongly symmetric data: the 4-layer EuroSAT model drops only 1.07% at 83% reduction. On weakly symmetric data (CIFAR-10), the pruned model can even gain 2.6 points, but our analysis attributes this to relaxing a mismatched equivariance constraint rather than to pruning itself; the value of pruning over from-scratch training scales with the data’s symmetry strength.
Förderung
- King Faisal University
Zitieren
APA
Alnemari, M., & Al-Omair, O. M. (2026). Symmetry-Preserving Pruning of Group Equivariant Convolutional Networks via Representation Theory. Symmetry, 18(6), 983. https://doi.org/10.3390/sym18060983
Vancouver
Alnemari M, Al-Omair OM. Symmetry-Preserving Pruning of Group Equivariant Convolutional Networks via Representation Theory. Symmetry. 2026;18(6):983. doi:10.3390/sym18060983
BibTeX
@article{Alnemari2026,
author = {Alnemari, Mohammed and Al-Omair, Osamah M.},
title = {Symmetry-Preserving Pruning of Group Equivariant Convolutional Networks via Representation Theory},
journal = {Symmetry},
year = {2026},
volume = {18},
number = {6},
pages = {983},
doi = {10.3390/sym18060983}
}
EUVIMED