A group of AI scientists at Universitat Politècnica de València, in Spain, has actually located that as prominent LLMs (Big Language Versions) enlarge and much more advanced, they come to be much less most likely to confess to a customer that they do not understand a solution.
发布者:Dr.Durant,转转请注明出处:https://robotalks.cn/as-llms-grow-bigger-theyre-more-likely-to-give-wrong-answers-than-admit-ignorance/