Notes · Series

Inside LLMs

What happens between the prompt and the answer: practical notes on reading a model’s internal representations, and what they can and cannot tell us.