ResearchBenchmarksFine-grained Query Answering on Clinical Narratives Item-level (test)Follow60.3ROUGE-LLENS11.73224.34136.9549.559Dec 28, 2025Evaluation ResultsMethodMethodLinksROUGE-LBERTScoreCoveragePresence AlignmentLENSevaluation_level=item-...evaluation_level=item-level, parameters=14B, prompting=zero-shot2025.1260.383.2—73.2TS-Textevaluation_level=item-...evaluation_level=item-level, parameters=14B, prompting=few-shot2025.1214.260.4—66.5TS-Imageevaluation_level=item-...evaluation_level=item-level, parameters=32B, prompting=few-shot2025.1213.660.2—68.7