AI Paper Atlas

Research map / Architectures and efficiency / RNNs and normalization

Abstraction Mechanisms Predict Generalization in Deep Neural Networks

Alex Gain, Hava Siegelmann

ICML 2020 · Architectures and efficiency › RNNs and normalization · approach: Mechanistic circuit interpretability · 3 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
Use CNA to predict test accuracy across architectures and datasetsMLP, VGG-18, ResNet-18, and ResNet-101 · ImageNet, CIFAR-10, CIFAR-100, SVHN, MNIST, and Fashion-MNISTtest accuracyImproved
Use CNA-Margin rather than norm-based metrics to predict the generalization gapDNNs across the evaluated architectures · Gaussian noise datasetPearson correlation with generalization gapImproved
Use CNA-Margin rather than competing metrics to predict the generalization gapDNNs across the evaluated architectures · Datasets with 10% to 50% shuffled training labelsPearson correlation with generalization gapImproved
Evaluate test error on inputs with higher information complexityMLP · MNISTtest error: up to 16 times larger test errorHurt
Train the MLP on MNIST and track CNA over trainingMLP · MNISTtraining lossNo 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