LLM's Are Great! But What is actually going on under the hood? (Part 1). (opens in new tab)

A clear, code-first guide to transformer blocks, attention, KV cache, and why LLM costs scale with context length.