An interactive visualization of a small neural network that learns the XOR function. Click input neurons to toggle values, adjust weights and biases, and watch activations propagate through the network in real time.
Live demo: https://loehx.github.io/xor-neural-network/
- Visual feed-forward network with input, hidden, and output layers
- Click input neurons to set 0 or 1
- Click synapse weights and neuron biases to edit them
- Preconfigured XOR weights loaded on startup
- State persisted in the browser via
localStorage
Open index.html in a web browser, or serve the project with any static file server:
python3 -m http.server 8000Then visit http://localhost:8000.
MIT © Alexander Löhn