Are Larger Pretrained Language Models Uniformly Better? Comparing Performance at the Instance Level.
R. Zhong, D. Ghosh, D. Klein, und J. Steinhardt. ACL/IJCNLP (Findings), Volume ACL/IJCNLP 2021 von Findings of ACL, Seite 3813-3827. Association for Computational Linguistics, (2021)