Building a Mini Network

The Story

One neuron can only do so much — it draws a single line in the sand. But wire a few together, and something magical happens: emergence. The network can learn patterns that no single neuron could represent.

The Layer

A layer is just a collection of neurons that all receive the same inputs, but each has its own weights and bias. If we have 2 inputs and 3 neurons in a layer, we need 6 weights (2 per neuron) and 3 biases.

def layer(inputs, all_weights, all_biases):
    return [neuron(inputs, w, b) for w, b in zip(all_weights, all_biases)]

The Network

Stack two layers and you have a network. The output of the first layer becomes the input to the second. That's it — a feedforward neural network.

# Layer 1: 2 inputs -> 3 neurons
hidden = layer(inputs, weights1, biases1)

# Layer 2: 3 inputs (from layer 1) -> 1 neuron
output = layer(hidden, weights2, biases2)

Why This Matters

A single neuron draws a line. A layer draws many lines. A network of layers can draw curves, contours, and complex decision boundaries. This is the power of composition.