AI Paper Atlas

Research map / Architectures and efficiency / CNN design

Platonic Transformers: A Solid Choice For Equivariance

Mohammad Mohaiminul Islam, Rishabh Anand, David R Wessels, Friso de Kruiff, Thijs P. Kuipers, Rex Ying, Clara I. Sánchez, Sharvaree Vadgama, Georg Bökman, Erik J Bekkers

ICML 2026 · Architectures and efficiency › CNN design · approach: Group symmetry equivariant networks · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add 2D rotational group weight-sharing (D3) to the ViT (ablation)ViT with RoPE · CIFAR-10accuracy (attention): 91.65 -> 92.78Improved
use a larger symmetry group (C6/D3) with fewer parameters (ablation)Platonic Transformer · CIFAR-10accuracy: 14.2M params vs 85.1M for baselineImproved
keep softmax attention instead of linear dynamic convolution (ablation)Platonic Transformer · CIFAR-10accuracy: +5.31 percentage points averaged over groupsImproved
add Tetrahedron group weight-sharing (ablation)ViT with RoPE · ScanObjectNNoverall accuracy (attention): 83.68 -> 86.11Improved
keep softmax attention instead of linear dynamic convolution (ablation)Platonic Transformer · ScanObjectNNoverall accuracy: +6.26 percentage points averaged over groupsImproved
use equivariant Tetrahedron group instead of trivial group (ablation)DiP diffusion transformer · QM9 generationmolecule stability: 91.54 -> 95.26Improved
use equivariant Tetrahedron group instead of trivial group (ablation)DiP diffusion transformer · QM9 generationvalidity (%): 97.09 -> 98.43Improved
use Tetrahedron group instead of trivial group (ablation)Platonic Transformer · ProteinMD backbone levelforce MSE: 2.11 -> 1.80Improved
use Tetrahedron group instead of trivial group (ablation)Platonic Transformer · ProteinMD atom levelforce MSE: 2.60 -> 2.36Improved
increase symmetry group from Tetrahedron to Octahedron at halved parameters (30M) (ablation)Platonic Transformer · OMol25 4M (20 epochs)force MAE: 11.15 -> 12.28Hurt
scale Octahedron model back to 60M parameters (ablation)Platonic Transformer Octahedron · OMol25 4M (20 epochs)force MAE: 12.28 -> 10.51Improved
train Tetrahedron model longer (20 to 80 epochs)Platonic Transformer Tetrahedron · OMol25 4Mforce MAE: 11.15 -> 9.33Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy