Research map / Architectures and efficiency / CNN design
Co-Attentive Equivariant Neural Networks: Focusing Equivariance On Transformations Co-Occurring in Data
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 |
|---|---|---|---|
| Add co-attentive equivariant mappings | p4-CNN · Rotated MNIST | Test Error (%): 2.28 ± 0.0004 -> 2.06 ± 0.0429 | Helped |
| Add co-attentive equivariant mappings | DREN · Rotated MNIST | Test Error (%): 1.78 (1.99) -> 1.674 | Helped |
| Add co-attentive equivariant mappings | DRENMaxPool · Rotated MNIST | Test Error (%): 1.56 (1.60) -> 1.34 | Helped |
| Add co-attentive equivariant mappings | p4-All-CNN · CIFAR-10 | Test Error (%): 8.84 -> 7.68 | Helped |
| Add co-attentive equivariant mappings | p4m-All-CNN · CIFAR-10 | Test Error (%): 7.59 -> 6.92 | Helped |
| Add co-attentive equivariant mappings | p4m-ResNet44 · CIFAR-10 | Test Error (%): 6.46 -> 9.12 | Hurt |
| Add co-attentive equivariant mappings | r-NINx4 · CIFAR-10 | Test Error (%): 14.96 -> 13.67 | Helped |
| Add co-attentive equivariant mappings | r-ResNet20x4 · CIFAR-10 | Test Error (%): 12.31 -> 11.32 | Helped |
Closest papers by meaning
- A Probabilistic Approach to Learning the Degree of Equivariance in Steerable CNNs (ICML 2024 · similarity 0.89)
- A Characterization Theorem for Equivariant Networks with Point-wise Activations (ICLR 2024 · similarity 0.89)
- Flow Equivariant Recurrent Neural Networks (NeurIPS 2025 · similarity 0.88)
- Affine Equivariant Networks Based on Differential Invariants (CVPR 2024 · similarity 0.88)
- Neural Isometries: Taming Transformations for Equivariant ML (NeurIPS 2024 · similarity 0.88)
- A Tale of Two Symmetries: Exploring the Loss Landscape of Equivariant Models (NeurIPS 2025 · similarity 0.88)
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.