1 paper
Xinrui Ruan, Zhenyu Zhao, Waverly Wei +4
Reliable generative AI models critically rely on expert human annotations to evaluate output quality, yet these "gold" labels are expensive to collect and limited in quantity. Orga…