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