Can't find an article on neural networks?

Question

Habr

Recently I read an article that described the fact that the more internal layers are added to the network, the worse it works, and the ways to solve this problem.
It turns out that I forgot to bookmark it and now I can’t find it, please discard, who remembers
PS any literature on this topic is also welcome, especially where there is more code and less formulas

Answer 1

Frankly speaking, it works just as well, it's just that the network has more parameters to configure and is more flexible.
There is such a thing as retraining, so the larger the network, the easier it is to retrain.