Writing
Notes on probability, stochastic processes and machine learning. Some posts are interactive, so you can move the parameters yourself and watch the mathematics respond.
Brownian motion depends on time and on chance. Freeze one, vary the other, and watch sample paths, random variables and increments appear. Built for my MATH 448 directed study in stochastic calculus.
Every Gaussian comes from one formula controlled by two numbers. Drag the mean and standard deviation, and see what each term of the density is responsible for.
A look at the Center for AI Safety's paper on AI loyalty subversion, and where I'm not fully convinced the Cold War deterrence analogy holds.
A LightGBM model that predicts a New York City listing's reviews per month from price, location and room type, and what its feature importances reveal about why some listings take off.