← Back to library

LLM basics

How an LLM Generates a Response

Follow a user prompt from tokenization through attention and decoding to a final answer.