Examination of whether LLMs can maintain consistency over extended multiple text generation for 10 medical personas. 5 novel plausibility metrics proposed, and an ontology of common LLM errors.
nlp ai medical question-answering bart mmr gen llm flant5 llm-inference llama2 llama2-7b nlp-medical-records maximal-marginal-relevance
-
Updated
Jul 12, 2024 - Python