AI Paper Atlas

Research map / Architectures and efficiency / RNNs and normalization

Characterizing the Discrete Geometry of ReLU Networks

Blake B. Gaines, Jinbo Bi

ICLR 2026 · Architectures and efficiency › RNNs and normalization · approach: Width and depth of ReLU networks · 1 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
increase depth from 1 to 4 at width 4 (ablation)ReLU network · Synthetic Gaussians, input dim 4average degree: 4.00±0.00 -> 6.17±0.12Improved
increase width from 4 to 16 at depth 1 (ablation)ReLU network · Synthetic Gaussians, input dim 4average degree: 4.00±0.00 -> 7.32±0.00Improved
increase depth from 1 to 4 at width 16 (ablation)ReLU network · Synthetic Gaussians, input dim 4average degree: 7.32±0.00 -> 7.85±0.03Improved
increase input dimension from 4 to 5 at depth 4, width 16 (ablation)ReLU network · Synthetic Gaussiansaverage degree: 7.85±0.03 -> 9.80±0.03Improved
increase depth from 1 to 4 at width 16 (ablation)ReLU network · Synthetic Gaussians, input dim 4graph diameter: 22.50±0.35 -> 76.35±4.56Hurt
increase width from 4 to 16 at depth 1 (ablation)ReLU network · Synthetic Gaussians, input dim 4graph diameter: 5.50±0.00 -> 22.50±0.35Hurt
change input dimension from 4 to 5 at depth 1, width 8 (ablation)ReLU network · Synthetic Gaussiansgraph diameter: 10.60±0.42 -> 10.75±0.26No clear effect
change input dimension from 4 to 5 at depth 1, width 16 (ablation)ReLU network · Synthetic Gaussiansgraph diameter: 22.50±0.35 -> 23.17±0.41No clear effect
compare regions containing training data vs regions without datatrained ReLU networks · MNIST, CIFAR10, California Housingneighbor countImproved
compare proportion of unbounded data-containing regions, classification vs regressiontrained ReLU networks · MNIST, CIFAR10 vs California Housingfraction of unbounded regionsMixed

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