Books
- Reinforcement Learning from Human Feedback
How language models are tuned after pretraining: instruction tuning, reward models, policy-gradient methods, direct alignment, and the preference data behind them. Each chapter pairs the ideas with the practical choices that make them work.
- Deep Learning
The standard reference for the foundations of neural networks: the linear algebra, probability and optimization they rest on, the practical networks used today, and the research ideas behind generative models.