Open-source lecture notes, concept wikis, flashcards and exercises for MSc Artificial Intelligence at the University of Amsterdam.
Current courses
Philosophy of Science
Logical empiricism, Popper and Lakatos, Kuhn, under-determination, explanation and understanding, and the scientific realism debate. Exam on Mon 19 Oct 2026.
→ Course home · Flashcards · Overview
Multilingual NLP
Writing systems, morphology, subword segmentation, static and contextual embeddings, and crosslingual transfer. Exam on Thu 22 Oct 2026.
→ Course home · Flashcards · Overview
Every lecture note in these two courses ends with a set of flashcards. Each course’s flashcards page lets you study the whole course, one lecture, or any mix of lectures.
Past courses
Reinforcement Learning
Sutton & Barto + UvA lecture series. Covers MDPs, dynamic programming, Monte Carlo, TD learning, function approximation, and deep RL.
→ Course home · Overview
Information Retrieval
Classical and neural retrieval models. BM25, language models, evaluation metrics, dense retrieval, learned sparse retrieval, and generative IR.
→ Course home · Overview
Recommender Systems
A short UvA lecture series (no textbook). Collaborative filtering and matrix factorization, beyond-accuracy evaluation (diversity, novelty, fairness), sequential models (GRU4Rec, SASRec, BERT4Rec), and generative recommendation (semantic IDs, RQ-VAE, TIGER, LRMs).
→ Course home · Overview
Concepts
The Concepts folder is a flat wiki of about 275 interconnected notes spanning all the courses. Each concept links to the lectures where it appears and to related concepts. Use the graph view or backlinks to explore connections.
How to use these notes
- Flashcards: open a course’s flashcards page, pick the whole course or single lectures, and press Study
- Search (Ctrl/Cmd+K) to find any topic
- Graph view to explore how concepts connect
- Backlinks at the bottom of each page show what links here
- All formulas are in LaTeX, all definitions use callouts
About
Built by Stanisław Wasilewski using Obsidian + Quartz. Notes are designed to fully substitute lecture attendance: every formula explained term-by-term, every figure reproduced.
Source: GitHub