# Prior knowledge established: Programming background, no ML

User is a programmer comfortable with Python and APIs, starting to learn LLM internals. No formal ML/DL/math background. Learning goal is practical — understanding LLMs to build better AI applications, not to become a researcher. Teaching should use programming analogies (functions, data structures, maps) rather than mathematical formalism. Chinese explanations preferred, English technical terms alongside. English reading ability is fine for source materials.

**Implications**: Build every concept from programming intuition first. Introduce math only when absolutely necessary, and always with concrete examples. Every lesson must connect back to "what this means for your code."