1 paper
Minhui Zhang, Prahar Ijner, Yoav Wald +1
Large Language Models (LLMs) often exhibit systematic errors on specific subsets of data, known as error slices. For instance, a slice can correspond to a certain demographic, wher…