Introduction

Welcome to the RL Handbook. This is a guide to reinforcement learning, written to take the reader from the very basics of the field all the way to modern topics. The goal is to give a clear and consistent path through the main ideas of RL, with the math, code examples and the intuition presented side by side. It is written for researchers and students who are studying the field, and also serves as a reference for those already working in RL who want to revisit a specific topic. The handbook is open source and a living document: new chapters and updates will be added over time. I am a student researcher in RL, and this project started from the notes I wished I had when I was first learning the field.

This section, Introduction, is meant as a gentle starting point. The chapter "What is Reinforcement Learning?" introduces the agent–environment loop and the basic vocabulary of the field, and "Taxonomy of Methods" gives a map of the algorithm families that the rest of the handbook covers. If you already have some background in RL, you can skip these two chapters and go directly to Chapter 1, where the technical content begins.

Every chapter also has two small controls near the title. Copy Markdown copies the chapter source, so you can paste it into a language model and ask about a derivation, a code snippet, or a paragraph that feels unclear. Feedback opens a GitHub issue for that chapter. Corrections, questions, and contributions are very welcome; they help keep the handbook accurate and useful.