🧠 Two-Spiral Problem: Backpropagation, Loss Curve and Decision Boundary (2D)
A trainable multilayer perceptron on a flat canvas: decision boundary, network diagram with real weights, loss and accuracy curves and per-layer gradient sizes on the two-spiral problem.
⚙ Under the hood
The same trainable multilayer perceptron on a flat canvas: decision boundary, network diagram with weights, loss and accuracy curves and per-layer gradient sizes while it learns the two-spiral problem.
neural networkbackpropagationmultilayer perceptrondecision boundarygradient descentloss curvemachine learning
2D · HTML5 Canvas 2D · 60 FPS target · runs fully client-side, no install