Vendor-neutral fundamentals of working with large language models: tokens and sampling, prompting patterns, embeddings and vector search, retrieval-augmented generation, and the failure modes that bite in production.
You’re reading a free preview. Sign up to see every topic and question — and to play this deck as games.
The prompt shapes that reliably change output quality — and the ones that only look like they do.
80 total questions
Card #1
Zero-shot prompting
Card #2
Few-shot prompting
Card #3
Why must few-shot examples be formatted consistently?
1. Few-shot prompting means:
2. Chain-of-thought prompting helps most reliably on:
3. Wrapping user-supplied text in explicit delimiters primarily:
1. Providing worked input-output examples inside the prompt is called ___-shot prompting.
2. Asking for the task with instructions and no examples at all is ___-shot prompting.
3. Asking the model to show intermediate reasoning before answering is ___-of-thought prompting.
