Multi-Layer Neural Network

Adjust architecture, weights, biases, and training

Network Architecture

i
i
i

Input Values

Weight & Bias Controls

Training

i
i
i
i

Save/Load

3D Plot Settings

i
i
Click "Forward Pass" to see network output

Binary State Vectors

2D Plot
Polyhedral Decomposition
70
Colors show binary activation patterns. Drag to pan, scroll to zoom.
Dual Graph
Graph of adjacent regions
Training Loss History