ModelRefs / Active Prompting — AI Glossary

Active Prompting — AI Glossary

Selecting the most informative few-shot examples to annotate by measuring model uncertainty, then using human-annotated examples.

Overview

Diao et al. (2023) proposed selecting CoT few-shot examples by measuring disagreement across multiple model samples (high uncertainty = high informativeness), then having humans annotate those examples. Outperforms random few-shot selection across multiple reasoning benchmarks with the same annotation budget.

Reference details

Topicprompting
Last reviewed2026-06-24

Continue your research

Use these connected ModelRefs sections to compare alternatives, inspect implementation paths, and review the evidence and governance boundaries relevant to Active Prompting — AI Glossary.

Frequently asked questions

What is Active Prompting?

Selecting the most informative few-shot examples to annotate by measuring model uncertainty, then using human-annotated examples.

What concepts are related to Active Prompting?

Closely related concepts include few shot, chain of thought, calibration.