Rethinking Retrieval-Augmented Generation for Medicine: A Large-Scale, Systematic Expert Evaluation and Practical Insights
Open paper
Benchmark
Model
Rank
Results
medical-question-answering-on-medxpertqa-text
GPT-4o + Enhanced RAG (MedCPT k=16)
–
Accuracy: 35.4
medical-question-answering-on-medxpertqa-text
Llama-3.1-8B + Enhanced RAG (MedCPT k=16)
–
Accuracy: 21.8
Rank counts only results with a code link.