Teaching lab

A note from the workshop

A neuron you can play with

Four switches, a weighted calculation, and a cat used to explain an artificial neuron.

Four things to change

For an Ampli lesson in March 2024, I made a small interactive model of an artificial neuron. A fictional cat supplied the example. A learner could switch four inputs on or off and see the illustrated neuron's activation change.

The calculation combined the inputs with weights and a bias, then applied tanh. An SVG visualization showed the result. A change to an input had a visible consequence.

The example and my part

The cat analogy came from the lesson material. I built an interactive example around it so a learner could connect the controls, the calculation, and the result.

It was a simplified teaching prototype within my wider work on frontend lessons, programming projects, and later AI courses. Four switches were enough to compare different combinations against the same calculation.

A question to try

What do you expect to happen if one input changes? The model gives a way to make a prediction, try it, and inspect the difference. That is a concrete conversation an explanation can begin.

This project note was compiled with AI assistance from personal development records. It is a retrospective, rather than a post originally published at the time.

A few places to start