Are Larger Pretrained Language Models Uniformly Better? Comparing Performance at the Instance Level

33850

Back to Articles Dashboard

Knowledge Triples

Source Entity	Relationship	Target Entity
BERT Үлкен Material	comparesWith	BERT MINI OtherEntity
BERT MEDIUM OtherEntity	correlatesWith	BERT Үлкен Material
BERT MINI OtherEntity	correlatesWith	BERT MEDIUM OtherEntity
MNLI Task	evaluatedIn	BERT Үлкен Material
QQP Material	evaluatedIn	BERT Үлкен Material
SST 2 Material	evaluatedIn	BERT Үлкен Material