For each image passed through this neural network, the final output will have a calculated value in the 'cat' output neuron, and a calculated value in the 'dog' output neuron. The output neuron that received the highest score becomes the class prediction for the image used as input

In the animation, you can see neurons as different shades of color, meant to illustrate how the weights and biases of a neural network produce different neuron values, based on the input.