Neural Network Capacity Explorer

Train a small tabular MLP and watch training AUC climb with capacity while internal test AUC refuses to follow.

neural network
overfitting
capacity
generalization gap

Training AUC

0.833

Internal test AUC

0.802

External test AUC

0.744

Training loss

Generalization gap

Capacity sweep: hidden units vs. AUC

TrainingInternal test

Model comparison

Dataset	AUC
Training	0.833
Internal test	0.802
External test	0.744

AUC on internal and external test sets.

Capacity is not the bottleneck

With roughly a hundred training cases and eight tabular predictors, increasing hidden units drives training AUC toward 1.0 while internal test AUC stays flat or drops. This gap — the generalization gap — is the single most important diagnostic in deep learning. No combination of learning rate, epochs, or regularization manufactures signal that the data do not contain; deep networks earn their keep on raw voxels with thousands of studies, not on a handful of tabular features.

Preparing the simulation…