Tag
This paper introduces a sentence-level benchmark for classifying interpretive canons in legal documents, evaluating LLMs on a dataset from the German Federal Constitutional Court.