csinva.github.io

A personal website that hosts open lecture slides, research notes, cheat sheets, and blog posts about machine learning, statistics, and neuroscience.

• Author: Chandan — Senior Researcher at Microsoft Research, works on interpretable machine learning, PhD from UC Berkeley.
• Purpose: Publicly share his continuously-improved notes (taken since his PhD) and presentations.

What’s inside

• Slides (pres folder) — Presentation sources, written in Markdown and built with reveal-md, so they’re easy to edit and export. Includes ML slides from teaching Berkeley CS 189, AI slides from CS 188, and workshop slides on interpretability topics.
• Research overviews (_notes/research_ovws) — Summaries of recent papers in areas like interpretability, causal inference, transfer learning, uncertainty, deep-learning theory, complexity, scattering transform, and deep learning in neuroscience.
• Notes & cheat sheets (_notes + _notes/cheat_sheets) — Markdown notes for many courses/areas spanning computer science, statistics, and neuroscience (e.g., computational neuroscience, causal inference, classification, linear algebra, info theory, computer vision).
• Blog posts (_blog) — Posts on machine learning / statistics / neuroscience topics; selected examples include paper-writing tips (2023), a paper-title generator (2022), and a post about the imodels library (2022, on the Berkeley AI Research blog).
• Personal info — Link to his Google Scholar profile.

Tech details

• Built with Jekyll, hosted on GitHub Pages, using the Timeline theme, particles.js for animations, and Jupyter Book for some notes.
• The pres slides are generated with reveal-md.

Usage & license

• The README explicitly says “feel free to use openly” (no formal license file mentioned).
• Updates: you can follow the author on Twitter (@csinva) or star the repo.

— All statements above are drawn directly from the README; nothing else was assumed.