How Large Language Models Work: Transformers, Tokens, and Embeddings Explained

I've been building with LLMs for a while now, and I still remember the moment someone asked me "but how does it actually work?" and I realized I couldn't give a clean answer. I knew how to call the AP