Study notes

Notes from working through theory-heavy books and papers: derivations, worked exercises and interactive visualisations to build intuition. I used AI heavily to help me understand these texts. The derivations and explanations are mostly AI-written, in response to the questions I asked while studying, so they are not in my own words. They record my learning path rather than finished expert exposition, and may contain mistakes.

• Information geometry notes: chapter-by-chapter study notes on Amari’s Information Geometry and Its Applications, with derivations, runnable checks and interactive visualisations.
• A First Course in Machine Learning notes: notes, exercise solutions and interactive visualisations (Bayesian inference, Laplace approximation, MCMC, classification) for Rogers and Girolami’s textbook, with Python code.
• Theory-inclined paper notes: close readings of theory-heavy machine learning papers, one annotated page per paper.
• Bayesian deep learning notes (planned, nothing written yet): a path through Yarin Gal’s PhD thesis and Chin-Wei Huang’s Bayesian deep learning reading list.