Prerequisites & Notation

This chapter assumes familiarity with:

How LLMs work (Chapter 35): transformer architecture, pre-training, RLHF
Python fundamentals (Chapter 1): async, HTTP requests, JSON We cover the practical side: calling LLM APIs, designing effective prompts, implementing tool use, and building RAG systems.

Symbol	Meaning
$T_\text{in}$	Input token count
$T_\text{out}$	Output token count
$k$	Number of retrieved documents in RAG
$\mathbf{q}$	Query embedding vector
$\mathcal{D}$	Document corpus for retrieval