Notes

 neural arch vs researcher

https://x.com/shao__meng/status/2099281275370754227/photo/1 notes

reinventing ed as next token pred

https://www.interconnects.ai/p/open-source-ai-reading-list

https://deep-learning-alchemy.github.io/ from 9/28

https://x.com/gordic_aleksa/status/2099884333305860516 - "double descent can be understood as a phase transition between memorizing individual examples and learning reusable structure from the data"

https://web.stanford.edu/class/msande319/ -

MS&E 319: Efficient Generative Language Models - notes soon

https://spectrum.ieee.org/inference-hardware-revolution

https://github.com/wafer-ai/gpu-perf-engineering-resources

https://prove2.me/

https://alisawuffles.notion.site/alisa-s-book-of-llms

Confusion matrix

https://archerhume.com/posts/jevs-architecture-unmasked/