Projecting high-dimensional tensor/matrix/vect GPT–>ML

My logic meets hilbert spaces here: 3 general and a few specific solutions for DL and AI, including ML visualization: either zoom in, look it in higher space, or zoom out and simplify it's space to yours. https://github.com/tambetvali/LaegnaAIMLBasics - machine learning intro of me, and I got many readers/cloners from here and like 4 stars (typically you guy use 2 computers and do not read online; that's the free stats from github: AIExperiments repo of mine brings some code for GPT) for AIBasics: but I rather found out it was very GPT related by expressing perceptron, basically GPT model is an architecture on perceptron. I am advocating the use of ML here. So all my news in one post if you dare to read comments :)