Andrej Karpathy’s “Deep Dive into LLMs like ChatGPT,” posted to his own channel in February 2025, is the long-form successor to his 2023 one-hour “Intro to Large Language Models” talk. This one runs about three and a half hours and follows a model from raw text to a finished assistant: pretraining data and tokenization, the next-token-predicting network, the supervised finetuning that turns a base model into a chatbot, and the reinforcement learning stage that shapes its behavior and reasoning. It is taught by someone who trained these systems at OpenAI and Tesla.