AI Paper Atlas

Research map / Architectures and efficiency / RNNs and normalization

Learning Neural Network Subspaces

Mitchell Wortsman, Maxwell C Horton, Carlos Guestrin, Ali Farhadi, Mohammad Rastegari

ICML 2021 · Architectures and efficiency › RNNs and normalization · approach: Efficient network compression and training · 14 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
Train a line subspace with cosine-similarity regularization (β > 0)Neural network lines · CIFAR-10, ImageNetaccuracyImproved
Use a layerwise sampling variant (ablation)Neural network subspaces · Image classification benchmarksmidpoint accuracyImproved
Use a layerwise sampling variant (ablation)Neural network subspaces · Image classification benchmarksensemble accuracyHurt
Increase the number of simplex endpoints (ablation)Simplex midpoint · CIFAR-10, Tiny-ImageNetaccuracyImproved
Use the simplex midpoint instead of a standard-trained modelSimplex · CIFAR-10 with label noiseaccuracyImproved
Use the simplex midpoint instead of a standard-trained modelSimplex · CIFAR-10expected calibration errorImproved
Train with multiple independent subspace samples per batch instead of one (ablation)Lines and curves · Image classification benchmarksaccuracyImproved
Add feature-similarity regularization between sampled subspace models (ablation)Lines and curves · Image classification benchmarksaccuracyImproved
Initialize line endpoints from the same shared weights instead of independent random initializations (ablation)Learned line · CIFAR-10, ImageNetendpoint distanceNo clear effect
Add cosine-similarity diversity regularization to SWA checkpoints (ablation)SWA · Image classification benchmarksperformanceNo clear effect
Use a simplex midpoint instead of SWALayerwise simplex · Image classification benchmarksaccuracyMixed
Use two models from opposing ends of a learned line instead of two independently trained modelsLearned line · Image classification benchmarksensemble accuracyNo clear effect

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