Generative AI Few-Shot Prompting
Use a few consistent demonstrations to explain an ambiguous format.
How it works
Few-shot prompting includes examples of the desired transformation. Each example should represent the actual task and follow the same rules as the instructions. Contradictory demonstrations can teach the wrong pattern. Keep evaluation cases separate from prompt examples so you do not mistake memorizing the supplied demonstrations for success on new inputs.
How to use this lesson
This is an authored prompt exercise with a human-written acceptance guide. It is not a captured provider response or a promise that a model will produce the target. No provider account, API, model weights or live inference was used for the course execution review. If you try it in an external service, use an authorised environment, check its actual access and data settings, and record the product/model and run date.
Bhylu does not run a generative model or send this prompt to an AI service. Copy and download preserve the text only. External tools may require accounts, paid access or compatible inputs; this course does not promise availability or prices.
Prompt exercise
Return one label: schedule, supplies or unknown.
Examples:
"Do I bring scissors?" -> supplies
"When do doors open?" -> schedule
"Who won the football match?" -> unknown
Now classify: "Are paper sheets included?"
Acceptance guide — not a model transcript
The reference label for the new message is supplies.
The examples show both in-domain labels and an out-of-domain refusal. The final input differs from all demonstrations. When revising this prompt, use a held-out set of other questions and inspect per-label errors rather than checking only the three examples already in context.