How LLMs Actually Work — Interview Cheat Sheet

A print-ready, 2-page A4 interview cheat sheet distilled from my LLM video.

Front: the entire mechanism in under 180 words — next-word prediction loop, tokens, training as trillions of fill-in-the-blank rounds, hundreds of billions of parameters, attention and the transformer (the T in GPT), and why hallucination falls out of the design. With the next-word-scores diagram ("The cat sat on the ___ — mat 62%").

Back: the 6 openers of every GenAI interview — what an LLM does in one sentence, what tokens are, how training works, what attention solves, why hallucination isn't a patchable bug, and why the same model that summarizes brilliantly cites sources badly — complete answers, plus 3 traps.

For GenAI-engineer and SDE interview prep. Instant download.