ChatGPT models excel in neurology exam, surpassing human student performance

Study shows two ChatGPT large language models (LLMs) outperformed human neurology students in an entrance exam, with one model achieving 85% accuracy, demonstrating their potential in clinical neurology and medical education.