Hi.

I am trying to get the output of the neural network layers (last second layer) in equation form; basically wanted to observe what type of basis functions the network has created in the process. I can get to show the output of the last second layer in the form of a plot; but was hoping to get it in mathematical form/equations.

Any leads will be appreciated. Thank you

You might find it helpful to work on some basic neural networks in spreadsheets to develop an intuition for what is going on behind the scenes.

Pytorch docs also can be helpful for understanding the function behind a given layer type. Take, for instance, the commonly used Linear layer:

Here we see an input vector, `x`

, is matrix multiplied by a weight matrix, `A`

(that is transposed, hence the superscript `T`

), and then a vector of biases added, `b`

.

https://pytorch.org/docs/stable/generated/torch.nn.Linear.html