Masking self-attention to syntactic ancestors and averaging it with BERT attention improves SQuAD 2.0 exact match from 84.1 to 85.1 and RACE accuracy from 72.6 to 74.2.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2019 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
SG-Net: Syntax-Guided Machine Reading Comprehension
Masking self-attention to syntactic ancestors and averaging it with BERT attention improves SQuAD 2.0 exact match from 84.1 to 85.1 and RACE accuracy from 72.6 to 74.2.