Mohamed Tarek

There is a decent literature review in one of my paper drafts, https://arxiv.org/abs/2103.04594 which will most likely be published at SMO after I address some reviewers' comments.

Can you post an MWE?

Can you post an MWE?

Thanks for the summary. I will respond to each point below. 1. Regarding `ADTypes.jl`, `AbstractDifferentiation.jl` predates that package. But we can of course support their AD types with an extension...

Sorry I didn't see your comment and opened my own PR @sethaxen :) #85

Kind of. Currently the function returned by `value_and_pullback_function` takes a single input (co-tangents) and returns the function value and the pulled back co-tangents. So the value is returned in a...

it's not a show-stopper right now as most users won't even use this function

I would like to add a "batch" version of Zygote as a backend which falls back on Zygote except for `jacobian` where the pullback is called with all the bases...

And a `SparseDiffTools` backend to optimise for sparsity structure

>or just something that sometimes works? Something that sometimes works. The goal is to make it easy to define a sparse Jacobian in a rrule and then get it back...