Course on LLMs

The goal of this course is two-fold: it acts as a one-shop-stop for me to be able to go back and quickly revise topics regarding LLMs and, hopefully, it provides helpful for other fellow researchers and AI-safety enthusiasts.

This course will be math-focused. Being mathematically precise is a double-edged sword, the more precise you are, the harder the content is to parse and the harder I find it to come up with intuition for novel research directions. Yet, mathematical rigour is fundamental for correctness - Here I will try to strike a good balance. Eventually I will also add some code snippets (most likely using Pytorch).