My Summer List of Core AI Readings... (Part Deux!)
Wow, such an amazing response from so many people on other readings, I needed to make a second part!
Thank you all so much for the feedback, suggestions, critiques and memes off of my first post on AI readings! Here’s a curated followup list of some other great resources
** also for disclosure, there are a few from for profit organizations, products, etc. Do not consider my inclusion here as an endorsement, paid or otherwise **
Vector Databases: First a good introduction here, one of many good use cases here & a great paper on understanding sparse vs. dense vector search here.
LoRA: To quote my old friend Andy, “You can’t leave out LoRA!” This is important to doing cheap transfer learning and fine tuning. No more spending $millions to train a model - just use an existing foundation model and fine tune it with a small set of additional training data and BAM!
RAG (retrieval augmented generation): Easily the best way to streamline the creation of intelligent natural language processing models. Introduction here, paper here.
Visualizing A Neural Machine Translation Model (Mechanics of Seq2seq Models With Attention): A sequence-to-sequence model is a model that takes a sequence of items (words, letters, features of an images…etc) and outputs another sequence of items. A fantastic primer with links to original papers is here. H/T to my friend Jason on this one
Transformer (part 2): Another solid paper here if you want more info.
And if you’re coming to this from a biological starting point, looking at the human brain and the role billions of years of evolution have taken one must include Logic of Chance. h/t to Jimmie for this one.