This is a real neural network learning in your browser! Each circle represents a neuron, and the lines show connections with weights. Watch as the network adjusts these weights to learn patterns in the data.
Forward Pass: Data flows from left to right through the network. Each neuron receives inputs, applies weights, adds bias, and uses an activation function.
Backward Pass: The network calculates its error and adjusts weights using backpropagation - sending error signals backward through the network.
🟢 Green connections: Positive weights (excitatory)
🔴 Red connections: Negative weights (inhibitory)
💫 Brightness: Activation strength or weight magnitude
⚡ Animation: Data or gradients flowing through the network
• Try different datasets to see how the network adapts
• Adjust the learning rate to control training speed
• Add more hidden layers for complex patterns
• Click on neurons to see detailed information