Tracking research paper changes
I recently came across an interesting thread on Twitter discussing a hypothetical scenario where research papers are published on GitHub and subsequent papers are diffs over the original paper. Information overload has been a real problem in ML with so many new papers coming every month.
