Work

A Rasch-based analysis comparing the performance of updated multimodal large language models and surgeons on the Japanese surgical specialist examination