An open, interactive introduction to reinforcement learning. AI-generated, human-curated, refined in the open.
Different goals, different content. Choose your path based on what you want to accomplish.
Learn the concepts
Progressive lessons from bandits to deep RL. Intuition, math, and code at your pace.
Deep dive research
Deep dives into influential RL papers, starting with GRPO. Context, insights, and critical analysis.
Solve real problems
Formulate real problems as RL: LLM fine-tuning, RLHF, elevator dispatch. End-to-end guides.
Scale and deploy
The engineering side of RL, starting with experiment tracking. More guides on the way.
Experiment hands-on
Interactive playgrounds to test algorithms. GridWorld runs in your browser, with more to come.
News and insights
Updates about rlbook.ai, guest posts, and insights on reinforcement learning from contributors.
Content is generated from carefully designed prompts, then reviewed and refined through community feedback.
Structure, examples, and learning objectives defined upfront
Content follows style guides and mathematical conventions
Every page shows its review status, from draft to verified — most content is still in review
Readers suggest improvements; content and prompts evolve
This is an open project. The content, prompts, and code are all on GitHub. Contributions of all kinds are welcome.
Typos, bugs, unclear explanations
Better prompts mean better content
Interactive visualizations and playgrounds
Help verify accuracy and quality
Created by Enes Bilgin,
author of "Mastering Reinforcement Learning with Python".
An experiment in AI-assisted education—creating high-quality learning resources through human-AI collaboration.