One way I try to keep myself thoroughly confused is by reading research papers.

I’ve recently been challenging myself to read a research paper a day. Over the past few weeks I have found some resources that I’d like to share. : This is of course the most straightforward tool. It is particularly useful in finding highly influential papers as the default rankings are based on number of citations. Unfortunately the results may not be accessible due to the occasional vomit-inducing paywall. : This is usually my first choice, (although it was the hardest for me to remember). Repository for open papers from computer science to astrophysics. They also have an Android app that allows you to download and store papers (which was great use while on a plane!)

citeseerx allows you to search through the Scientific Literature Digital Library (all papers are freely available) : A relatively new site but I cant recommend it enough! It ties a axriv paper with a freely viewable implementation in github! This makes it extremely convenient to learn exactly what an implementation of say, multi-dimensional neural network may look like. Once you realize that it all seems like magic, you can go and read the paper to pull you back into reality! (Or deeper into the rabbit hole…). In addition the site ranks of the most popular projects, giving you a better filter in comparison to the unranked behemoth of axriv papers.

These are the resources that stick most firmly in my mind. I’ll have to look through my massive list of bookmarks for more. If you know of other great resources. please let me know!