Results for “brownfield”

Turn model work into a production ML system with data contracts, reproducible training, quality gates, deployable artifacts, and monitoring.

Benchmarks zero-shot in-context learning on BIG-Bench Hard multiple-choice tasks, comparing self-generated demonstrations against direct prompting and chain-of-thought baselines, and reports accuracy.