Work

Response to correspondence regarding “Analysis of large-language model versus human performance for genetics questions”