Skip to main content
← Back to AI Foundations
Middle School Lab

Neural Net Visualizer

Build a neural network, set its inputs, and watch a real forward pass flow through every neuron.

Neural Network Visualizer

Build your own feed-forward neural network, adjust the inputs, then watch data flow layer-by-layer as the network computes its output. Every number you see is genuinely calculated — using the sigmoid activation function.

A neural network with 3 layers: an input layer of 3 neurons, one hidden layer of 4 neurons, and an output layer of 2 neurons. Activation values are shown as text inside each neuron circle.INPUTHIDDENOUTPUT
Positive weightNegative weightBrighter = higher activation
Input Values
Architecture

Hidden Layer4 neurons

All Activation Values — Text View

Input

N10.8000
N20.5000
N30.3000

Hidden

N10.8030
N20.4899
N30.7008
N40.6664

Output

N10.6081
N20.5838

What is a Forward Pass?

A neural network is made of layers of neurons. Each neuron receives numbers from the previous layer, multiplies each by a weight, adds them all up, then adds a bias. That total is called the weighted sum (z).

Then the neuron applies an activation function — here we use sigmoid: σ(z) = 1 / (1 + e^−z). Sigmoid squashes any number into the range (0, 1), which helps the network produce consistent outputs.

A forward pass is when data flows from left to right through all layers until we get the final output. Click any neuron to see its exact computation!