10. Neural networks and transformers
Build a plain-language picture of neural networks, attention, transformers, pretraining, instruction tuning, and alignment. You will know what LLMs are good at, where they are brittle, and why bigger is not always better.