measuring

Definition

Quantifying a property rather than improving it. A title beginning with "Measuring" usually introduces a benchmark, a metric, or an evaluation protocol, and the real argument is construct validity: does the number track the thing it names, or something correlated and easier to score? Such papers are the counterpart to improvement papers, which are only as trustworthy as the measurements they report against.

Explainers using this term

• LLM Evaluation from Scratch — Reading Benchmarks and the Contamination ProblemMeasuring Massive Multitask Language Understanding