Research map / Architectures and efficiency / RNNs and normalization
Geometric Algebra Transformer
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 changed | Where | Measured | Effect |
|---|---|---|---|
| use E(3)-equivariant geometric algebra Transformer instead of non-equivariant baselines (Transformer, MLP, GCA-GNN) | GATr · n-body dynamics | prediction error | Helped |
| GATr versus equivariant SEGNN and SE(3)-Transformer at small training sizes | GATr · n-body dynamics | prediction error vs training size | Helped |
| E(3) equivariance under spatial translation shift | GATr · n-body dynamics (translated data) | prediction error | Helped |
| GATr on randomly oriented arterial meshes versus baselines (Transformer, GEM-CNN, PointNet++) | GATr · Arterial wall-shear-stress | mean approximation error | Helped |
| canonicalize artery orientation for the Transformer baseline | Transformer · Arterial wall-shear-stress | mean approximation error | Helped |
| GATr as denoiser in Diffuser versus original Diffuser and Transformer backbone | GATr-Diffuser · Kuka block stacking | normalized reward | Helped |
| GATr versus Transformer in compute/memory scaling | GATr · synthetic scaling test | time and memory | Mixed |
Closest papers by meaning
- Lorentz-Equivariant Geometric Algebra Transformers for High-Energy Physics (NeurIPS 2024 · similarity 0.88)
- Geometric Algebra Planes: Convex Implicit Neural Volumes (ICML 2025 · similarity 0.83)
- DOCTR: Disentangled Object-Centric Transformer for Point Scene Understanding (AAAI 2024 · similarity 0.81)
- GLGENN: A Novel Parameter-Light Equivariant Neural Networks Architecture Based on Clifford Geometric Algebras (ICML 2025 · similarity 0.81)
- Composing Linear Layers from Irreducibles (NeurIPS 2025 · similarity 0.81)
- SE(3)-bi-equivariant Transformers for Point Cloud Assembly (NeurIPS 2024 · similarity 0.81)
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.