Pattern

Stratified sampling for human review

Choose which outputs a human reviews by stratifying the stream — oversampling low-confidence, rare-category, and high-stakes outputs — rather than sampling at random, so a fixed reviewer budget catches the most errors and the most consequential ones.

Classification Sampling the output stream for review by risk stratum instead of at random.

Full entry

The complete treatment — problem, how it works, when to use it, sources, and examples — lives in the volume below.

Browse the full pattern index or the catalog.

Search the catalog