A bridge-closure question needs a two-sentence explanation of why the span closed, not a three-word date. How should the models be ranked?
Select an answer to reveal the explanation.
Short Explanation
A two-sentence why-the-bridge-closed answer is not a three-word date. Match the metric to the shape: span exact match for extractive answers, a rubric plus optional overlap or a judge for long-form. Word count and classification accuracy do not rank those explanations.
Full Explanation
QA metrics must match the shape of the answer. Extractive exact match is for pointing at a span that already sits in the passage. A two-sentence why-the-bridge-closed answer needs a long-form protocol: a written rubric, and optionally overlap or a model judge. Word count and a classification accuracy headline do not rank those explanations.