How are these neurons arranged?

Question

Neural networks

There are a lot of demonstrations on the net, but I can’t find how they are arranged. We are talking about neurons that learn to move, for example:
Do I understand correctly that here the neural network makes a decision every time a frame is updated? On the basis of what data does it work, that is, what is the input? The coordinates of each node? Or angles between joints? Or something else? And what do we get as an output?