its mostly a mathematical equation based in connections and weights, thats why these things are notoriously difficult to debug :/.
Neural networks can be difficult to debug due to their complex, "black box" nature and the large number of parameters involved. Unlike traditional code, where you can step through each line, neural nets rely on interconnected layers and weights that are adjusted during training, making it harder to pinpoint the exact source of errors.
Lol, thats not how neural nets work.
its mostly a mathematical equation based in connections and weights, thats why these things are notoriously difficult to debug :/.
I know exactly how NNs work. I have built them myself - even doing the back propagation for training by hand (with a calculator).
and, as it happens, my if...then is exactly how NNs work
NNs are function modellers. It would be trivial to train an NN to respond with "I love you too, $user" to the input "I love you computer".