SUT30-000
An Introduction to the SUT30 thread
A month ago, I came across a repository on my GitHub feed called “sutkever-30-implementations”. As the name suggests, this repository contains detailed conceptual implementations of 30 foundational papers recommended by Ilya Sutkever, one of the most influential computer scientists of this decade.
This list of 30 papers, though a well-made collection of famous deep-learning ideas, initially felt like just another shortcut to “master” AI/ML. However, the way each paper from that list was covered in the repository, with simple yet detailed implementations of each concept, motivated me to look further.
Apart from the easy-to-learn “un-abstracted” implementations, which used nothing more than NumPy, the author also included multiple tracks for systematically learning those papers. The intention was not just to absorb information but to actively fill any existing knowledge gaps, which I also aim to do to develop a more intuitive and tangible connection with this field.
And so here I am, at the beginning of the first-ever project of this blog. I shall implement, code, and write my learnings from the 30 papers and blogs through this thread, for myself and everyone to refer to. I would refer to the same repository for any code and try to make the most out of it rather than simply copying and pasting it.
Let’s get started!
Hope you enjoy this!
Thread Navigation
SUT30-000: An Introduction to the SUT30 Thread.
References
P.S., I love this image of the Sunburst galaxy, whose light is being lensed into multiple images by gravitational lensing, creating the Sunburst Arc.
Amazing! Very excited to read the upcoming posts & learn along :)