Ishan Anand joins AJC and the Web Devs to demystify Large Language Models (LLMs) through the familiar (to some) tool of Excel.
We'll discuss the original GPT-2 architecture and learn how it can be implemented in a spreadsheet.
Along with learning the ins and outs of transformers, we'll hopefully answer the more important question of why anyone would want to do such a thing in the first place.
Links and Resources:
https://x.com/ianand
https://spreadsheets-are-all-you-need...
https://maven.com/spreadsheets-are-al...
https://www.latent.space/p/ai-engineer
https://thoughtvectors.substack.com/p...
https://blogs.rstudio.com/ai/posts/20...
https://statmodeling.stat.columbia.ed...
https://arxiv.org/pdf/2305.07759
https://jalammar.github.io/illustrate...
https://en.wikipedia.org/wiki/Frank_R...
https://news.ycombinator.com/item?id=...
https://www.transformer-circuits.pub/...
https://www.lesswrong.com/posts/tEPHG...