Neural Network Lab

A neural network learns to make predictions by training on data. Here you'll see the whole process step by step.

01
Train the network – The network starts with random weights. By training on hundreds of examples, it learns which connections matter.
02
Explore the calculations – Step through the network neuron by neuron. See how each neuron multiplies, sums, and activates.
03
Make your own predictions – Change the input values and see if the network answers correctly.

You start in 🧮 Calculation mode, where every neuron works with simple whole numbers. Once you understand how a neuron works, switch to one of the scenarios and let the network train itself.

AI-aktiviteter

Neural Network Lab

Follow the calculations step by step, train on data, and make predictions.

🇸🇪 Svenska
Scenario
Architecture
Input variables
Activation
Press Train the model to begin.
Training epochs
0

Concepts – click for an explanation

Input Neuron Weight Bias (+b) Activation Activation function Hidden layer Output Loss Forward pass Epoch Prediction

Training data

The first columns show real-world values, followed by the normalized (0–1) values the network sees.