Viewing a single comment thread. View all comments

fenixfunkXMD5a t1_irlgz2c wrote

I am new to ML but maybe it's just because simpler systems are better. If the auto encoder is forced to have a smaller output size, you're forcing it to be more decisive when learning

1