Evaluation of large language model performance on the Biomedical Language Understanding and Reasoning Benchmark

Benchmark Model Rank Results
named-entity-recognition-ner-on-ncbi-diseaseGPT-4–F1: 65.98
question-answering-on-bioasqGPT-4–Accuracy: 85.71
question-answering-on-blurbGPT-4–Accuracy: 80.56
question-answering-on-pubmedqaFlan-T5-XXL–Accuracy: 76.80