code2vec TL;DR

Recently, I made some slides about code2vec, a machine learning work using deeplearning to turn the abstract syntax tree of functions into continuous vectors... Basically, word2vec applied to programming languages. I think I managed to sum up decently the main points of the paper, so I decided to share it here.

While I'm at it, here's a very good list of papers about doing machine learning on code by the guys at source{d}.

Somewhat related, but more focused on the security side of things, is Fabian Yamaguchi's joern, which I highly recommend.

Enjoy my code2vec TL;DR!