A beginner-friendly walkthrough of how LLMs turn text into tokens, embeddings, attention patterns, probabilities, and generated responses.