Summaries and highlights of insightful papers in AI.
There are tons of cool papers out there. Yet, more often than not, academic papers are filled with redundacies and fancy terms/equations that are not always neccessary and increase the friction of understanding the main contribution of the work. Also for me, it is quite difficult to maintain the best ideas out of the paper in my memory for longer than a few seconds after I hit that close button and move on with my life. The goal of this repo is to find the minmal viable product (MVP) of cool innovative work in AI and neuroscience, package it in a concise format, along with my opinon, and document it here.
A summary is organized as follows:
Every paper exist for a reason; To solve some problem. Describe the problem in a couple of sentences.
The key premise of the paper. In simple words, what do does this paper do to solve this problem?
How does the proposed solution solve the problem.
The key challenges and techncial contributions introduced in this paper to make it work.
Cool story bro, but does it work? What does it work on? Does it really solve the problem?
My thoughts on the paper, how it can be extended, practical applications that I might think could be releveant or some ranting.
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent