The neuron brightness tells you the value of the neuron between 0-1 and the weights(lines connecting the neurons) are more green if there is a stronger direct relationship, and are red if there is a stronger inverse relationship, if there is little/no connection between neurons then that means there is little to no correlation between the 2
Sadly at the moment the backpropogation is not working im not sure why but I have already spent way too long on this so ill leave it for later👍
With this made, I can analyse large datasets to predict an outcome(with a bit of modification), for example image classification