Standard
BERT Busters: Outlier Dimensions That Disrupt Transformers. / Kovaleva, Olga; Kulshreshtha, Saurabh; Rogers, Anna; Rumshisky, Anna.
Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021. Online : Association for Computational Linguistics (ACL), 2021. s. 3392-3405.
Publikation: Bidrag til bog/antologi/rapport › Konferencebidrag i proceedings › Forskning › fagfællebedømt
Harvard
Kovaleva, O, Kulshreshtha, S
, Rogers, A & Rumshisky, A 2021,
BERT Busters: Outlier Dimensions That Disrupt Transformers. i
Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021. Association for Computational Linguistics (ACL), Online, s. 3392-3405.
https://doi.org/10.18653/v1/2021.findings-acl.300APA
Kovaleva, O., Kulshreshtha, S.
, Rogers, A., & Rumshisky, A. (2021).
BERT Busters: Outlier Dimensions That Disrupt Transformers. I
Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021 (s. 3392-3405). Association for Computational Linguistics (ACL).
https://doi.org/10.18653/v1/2021.findings-acl.300Vancouver
Kovaleva O, Kulshreshtha S
, Rogers A, Rumshisky A.
BERT Busters: Outlier Dimensions That Disrupt Transformers. I Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021. Online: Association for Computational Linguistics (ACL). 2021. s. 3392-3405
https://doi.org/10.18653/v1/2021.findings-acl.300Author
Kovaleva, Olga ; Kulshreshtha, Saurabh ; Rogers, Anna ; Rumshisky, Anna. / BERT Busters: Outlier Dimensions That Disrupt Transformers. Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021. Online : Association for Computational Linguistics (ACL), 2021. s. 3392-3405
Bibtex
@inproceedings{8ba37881dc604980993c523dc0a45912,
title = "BERT Busters: Outlier Dimensions That Disrupt Transformers",
abstract = "Multiple studies have shown that Transformers are remarkably robust to pruning. Contrary to this received wisdom, we demonstrate that pre-trained Transformer encoders are surprisingly fragile to the removal of a very small number of features in the layer outputs ($",
author = "Olga Kovaleva and Saurabh Kulshreshtha and Anna Rogers and Anna Rumshisky",
year = "2021",
month = aug,
day = "1",
doi = "10.18653/v1/2021.findings-acl.300",
language = "English",
pages = "3392--3405",
booktitle = "Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021",
publisher = "Association for Computational Linguistics (ACL)",
address = "United States",
}
RIS
TY - GEN
T1 - BERT Busters: Outlier Dimensions That Disrupt Transformers
AU - Kovaleva, Olga
AU - Kulshreshtha, Saurabh
AU - Rogers, Anna
AU - Rumshisky, Anna
PY - 2021/8/1
Y1 - 2021/8/1
N2 - Multiple studies have shown that Transformers are remarkably robust to pruning. Contrary to this received wisdom, we demonstrate that pre-trained Transformer encoders are surprisingly fragile to the removal of a very small number of features in the layer outputs ($
AB - Multiple studies have shown that Transformers are remarkably robust to pruning. Contrary to this received wisdom, we demonstrate that pre-trained Transformer encoders are surprisingly fragile to the removal of a very small number of features in the layer outputs ($
U2 - 10.18653/v1/2021.findings-acl.300
DO - 10.18653/v1/2021.findings-acl.300
M3 - Article in proceedings
SP - 3392
EP - 3405
BT - Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021
PB - Association for Computational Linguistics (ACL)
CY - Online
ER -
