How would I go about an implement operations that work in a batch like way, say you have the super obvious (Wx + b) but you need this to run over W0..Wk , x0..xk, b0..bk

i don’t anderstand what exactly you are asking about. any specific example?

I'll dig around a bit, I'm not very familiar to neanderthal yet, I suspect I'm looking for something similar or having the same properties as a tensor along dimension https://deeplearning4j.org/docs/latest/nd4j-overview#getsettensor , however I'm not sure the same thing applies

is there any support for 3d/4d arrays?