Key notes, papers to read

While watching the ARENA Lecture 1 on Transformers, I found a link to Andrej Karpathy’s video titled “Let’s build the GPT Tokenizer.”

https://www.alignmentforum.org/posts/jP9KDyMkchuv6tHwm/how-to-become-a-mechanistic-interpretability-researcher

https://www.neelnanda.io/mechanistic-interpretability/prereqs

https://www.alignmentforum.org/posts/NfFST5Mio7BCAQHPA/an-extremely-opinionated-annotated-list-of-my-favourite

https://transformer-circuits.pub/2021/framework/index.html

[Intepretability in the Wild - Video walkthrough](https://www.youtube.com/watch?v=gzwj0jWbvbo Mech interp is not pre-paradigmatic by Lee Sharkey