But I was thinking that it can be interesting to show people on video what the Neural Network think when we send it like "a cat".
So i have made a static function called reverse to make a new neural network with everything in a revert mode PLUS add a bias at the input layer but set by default to 0 to avoid conflict.
So it makes the new (invert) neural network works exactly like the standard one.
I know it's not usefull ! (Really)
But I was thinking that it can be interesting to show people on video what the Neural Network think when we send it like "a cat".
So i have made a static function called reverse to make a new neural network with everything in a revert mode PLUS add a bias at the input layer but set by default to 0 to avoid conflict. So it makes the new (invert) neural network works exactly like the standard one.