# Can I extract symbolic expression of neural network loss function?

Once we create a neural network with NetChain in Mathematica, can we extract the loss function in the symbolic form for Mathematica to play with symbolic manipulations using Mathematica's built-in functions?

Edit: I would also extend the question to symbolic gradients.

Thanks.

• Unfortunately I don't think this is possible yet. In future versions of the language, there will be functionality called NetFunction that will hopefully allow this. – Carl Lange May 1 '19 at 6:48