Selection-based prompting with synthesized candidates unlocks high-performance clinical abbreviation disambiguation via open-source LLMs
{{output}}
Objective: Clinical abbreviations introduce semantic ambiguity that hinders automated understanding in healthcare informatics. While generative large language models (LLMs) show promise, direct generation often lacks clinical pre... ...
