About This Document
- sl:arxiv_author :
- sl:arxiv_firstAuthor : Kevin Clark
- sl:arxiv_num : 1906.04341
- sl:arxiv_published : 2019-06-11T01:31:41Z
- sl:arxiv_summary : Large pre-trained neural networks such as BERT have had great recent success
in NLP, motivating a growing body of research investigating what aspects of
language they are able to learn from unlabeled data. Most recent analysis has
focused on model outputs (e.g., language model surprisal) or internal vector
representations (e.g., probing classifiers). Complementary to these works, we
propose methods for analyzing the attention mechanisms of pre-trained models
and apply them to BERT. BERT's attention heads exhibit patterns such as
attending to delimiter tokens, specific positional offsets, or broadly
attending over the whole sentence, with heads in the same layer often
exhibiting similar behaviors. We further show that certain attention heads
correspond well to linguistic notions of syntax and coreference. For example,
we find heads that attend to the direct objects of verbs, determiners of nouns,
objects of prepositions, and coreferent mentions with remarkably high accuracy.
Lastly, we propose an attention-based probing classifier and use it to further
demonstrate that substantial syntactic information is captured in BERT's
attention.@en
- sl:arxiv_title : What Does BERT Look At? An Analysis of BERT's Attention@en
- sl:arxiv_updated : 2019-06-11T01:31:41Z
- sl:bookmarkOf : https://arxiv.org/abs/1906.04341
- sl:creationDate : 2019-06-21
- sl:creationTime : 2019-06-21T21:49:32Z
Documents with similar tags (experimental)