##### Understanding Generative AI (Stable Diffusion) as Galton Board

Understanding Generative AI in particular 'Stable Diffusion' as 'Galton Board' focusing on core mathematical ideas. No prerequisite knowledge required.

Casual essays and thoughts on physics, mathematics and computing

Subscribe or
[enable JavaScript to see contact information]

Understanding Generative AI in particular 'Stable Diffusion' as 'Galton Board' focusing on core mathematical ideas. No prerequisite knowledge required.

Introduction to Roger Penrose's (2020 Nobel Prize winner in Physics) graphical notation as applied to vector calculus, and an example application to a classical physics problem - deriving wave equation from Maxwell's equations. Prerequisite - familiarity with vector calculus

Understanding Generative AI - the Continuous Normalizing Flows and Flow Matching model. In particular, we will take a lightly flavored physics angle of fluid dynamics. Prerequisites - calculus, and basic understanding of generative AI as such

Nontechnical story on whales, highways and solutions

Understanding how to speed up the decoding process of Large Language Models by speculative sampling. Prerequisites - basic understanding of how LLMs function.

Experimental evidence that the following two things are not true simultaneously - things have clear existence and things interact only adjacently. Experiment for likes of which the 2022 Nobel Prize in physics was awarded. No prerequisite knowledge required

Approaching heart of computing through visualized lambda calculus. No prerequisite knowledge required

Re-inventing special relativity using simplicity and elegance as guiding principles. No prerequisite knowledge required