通常用于评估两个分布之间的相似度,生成模型的KID就是基于该值 <img src="https://res.cloudinary.com/montaigne-io/image/upload/v1726071110/2E2FE47D-A00C-4948-A1B6-D48FA9A8FB1E.png" style="background-color:initial;max-width:min(100%,1520px);max-height:min(1014px);;background-image:url(https://res.cloudinary.com/montaigne-io/image/upload/v1726071110/2E2FE47D-A00C-4948-A1B6-D48FA9A8FB1E.png);height:auto;width:100%;object-fit:cover;background-size:cover;display:block;" width="1520" height="1014"> 参考链接:<a href="https://blog.csdn.net/tunhuzhuang1836/article/details/78058184">https://blog.csdn.net/tunhuzhuang1836/article/details/78058184</a>