Atiyo Ghosh

Also related is their follow up work which uses neuron-wise adaptive activations: https://arxiv.org/abs/1909.12228

Where is the best place for adaptive activations to live? Architectural aspects such as `FastDense` are in `DiffEqFlux` as far as I can tell. But does it make sense to...

To take this idea a bit further: the whole notion of efficient data acquisition for surrogates is essentially the same as active learning.From this perspective, one might dip into the...

Awesome. I'll plough on with this. It might take a little time while I familiarise myself with DiffEqJump. Edit: I haven't actually made much progress with this at all. Leaving...

Indeed mutating isn't supported by Zygote, which is used to calculate the gradients. It is supported in some other Julia AD packages which you might be able to use. However,...

Maybe ReverseDiff.jl will work?