Recurrent neural network from adder’s perspective: Carry-lookahead RNN

The recurrent network architecture is a widely used model in sequence modeling, but its serial dependency hinders the computation parallelization, which makes the operation inefficient. The same problem was encountered in serial adder at the early stage of digital electronics. In this paper, we disc...

Celý popis

Uložené v:
Podrobná bibliografia
Vydané v:Neural networks Ročník 144; s. 297 - 306
Hlavní autori: Jiang, Haowei, Qin, Feiwei, Cao, Jin, Peng, Yong, Shao, Yanli
Médium: Journal Article
Jazyk:English
Vydavateľské údaje: Elsevier Ltd 01.12.2021
Predmet:
ISSN:0893-6080, 1879-2782, 1879-2782
On-line prístup:Získať plný text
Tagy: Pridať tag
Žiadne tagy, Buďte prvý, kto otaguje tento záznam!
Abstract The recurrent network architecture is a widely used model in sequence modeling, but its serial dependency hinders the computation parallelization, which makes the operation inefficient. The same problem was encountered in serial adder at the early stage of digital electronics. In this paper, we discuss the similarities between recurrent neural network (RNN) and serial adder. Inspired by carry-lookahead adder, we introduce carry-lookahead module to RNN, which makes it possible for RNN to run in parallel. Then, we design the method of parallel RNN computation, and finally Carry-lookahead RNN (CL-RNN) is proposed. CL-RNN takes advantages in parallelism and flexible receptive field. Through a comprehensive set of tests, we verify that CL-RNN can perform better than existing typical RNNs in sequence modeling tasks which are specially designed for RNNs. Code and models are available at: https://github.com/WinnieJiangHW/Carry-lookahead_RNN.
AbstractList The recurrent network architecture is a widely used model in sequence modeling, but its serial dependency hinders the computation parallelization, which makes the operation inefficient. The same problem was encountered in serial adder at the early stage of digital electronics. In this paper, we discuss the similarities between recurrent neural network (RNN) and serial adder. Inspired by carry-lookahead adder, we introduce carry-lookahead module to RNN, which makes it possible for RNN to run in parallel. Then, we design the method of parallel RNN computation, and finally Carry-lookahead RNN (CL-RNN) is proposed. CL-RNN takes advantages in parallelism and flexible receptive field. Through a comprehensive set of tests, we verify that CL-RNN can perform better than existing typical RNNs in sequence modeling tasks which are specially designed for RNNs. Code and models are available at: https://github.com/WinnieJiangHW/Carry-lookahead_RNN.The recurrent network architecture is a widely used model in sequence modeling, but its serial dependency hinders the computation parallelization, which makes the operation inefficient. The same problem was encountered in serial adder at the early stage of digital electronics. In this paper, we discuss the similarities between recurrent neural network (RNN) and serial adder. Inspired by carry-lookahead adder, we introduce carry-lookahead module to RNN, which makes it possible for RNN to run in parallel. Then, we design the method of parallel RNN computation, and finally Carry-lookahead RNN (CL-RNN) is proposed. CL-RNN takes advantages in parallelism and flexible receptive field. Through a comprehensive set of tests, we verify that CL-RNN can perform better than existing typical RNNs in sequence modeling tasks which are specially designed for RNNs. Code and models are available at: https://github.com/WinnieJiangHW/Carry-lookahead_RNN.
The recurrent network architecture is a widely used model in sequence modeling, but its serial dependency hinders the computation parallelization, which makes the operation inefficient. The same problem was encountered in serial adder at the early stage of digital electronics. In this paper, we discuss the similarities between recurrent neural network (RNN) and serial adder. Inspired by carry-lookahead adder, we introduce carry-lookahead module to RNN, which makes it possible for RNN to run in parallel. Then, we design the method of parallel RNN computation, and finally Carry-lookahead RNN (CL-RNN) is proposed. CL-RNN takes advantages in parallelism and flexible receptive field. Through a comprehensive set of tests, we verify that CL-RNN can perform better than existing typical RNNs in sequence modeling tasks which are specially designed for RNNs. Code and models are available at: https://github.com/WinnieJiangHW/Carry-lookahead_RNN.
Author Shao, Yanli
Qin, Feiwei
Peng, Yong
Jiang, Haowei
Cao, Jin
Author_xml – sequence: 1
  givenname: Haowei
  orcidid: 0000-0002-5619-5252
  surname: Jiang
  fullname: Jiang, Haowei
  email: dsfdsf@hdu.edu.cn
  organization: School of Computer Science and Technology, Hangzhou Dianzi University, China
– sequence: 2
  givenname: Feiwei
  surname: Qin
  fullname: Qin, Feiwei
  email: qinfeiwei@hdu.edu.cn
  organization: School of Computer Science and Technology, Hangzhou Dianzi University, China
– sequence: 3
  givenname: Jin
  orcidid: 0000-0002-9880-0342
  surname: Cao
  fullname: Cao, Jin
  email: jcao21@jhu.edu
  organization: Whiting School of Engineering, Johns Hopkins University, USA
– sequence: 4
  givenname: Yong
  orcidid: 0000-0003-1208-972X
  surname: Peng
  fullname: Peng, Yong
  email: yongpeng@hdu.edu.cn
  organization: School of Computer Science and Technology, Hangzhou Dianzi University, China
– sequence: 5
  givenname: Yanli
  orcidid: 0000-0002-3238-5430
  surname: Shao
  fullname: Shao, Yanli
  email: shaoyanli@hdu.edu.cn
  organization: School of Computer Science and Technology, Hangzhou Dianzi University, China
BookMark eNqFkM1KAzEUhYMo2Kpv4GKWbma8yfxlXAhS1AqiILoOaXIHU6eTepNW3Pkavp5P4pS6cqGrsznfgfON2W7ve2TsmEPGgVen86zHVY8xEyB4BjKDXOywEZd1k4pail02AtnkaQUS9tk4hDkAVLLIR2z6gGZFhH1Mhg3S3RDxzdNL0pJfJNpapK-Pz5AskcISTXRrPEsmmug97bx_0c-obfJwd3fI9lrdBTz6yQP2dHX5OJmmt_fXN5OL29TkeRNTXhYzAFM0Vc3rFgrBQcykldWMt20DvNFosTYSmqrUra2KAtqytjOh85KjxvyAnWx3l-RfVxiiWrhgsOt0j34VlCjrcvjGJQzVs23VkA-BsFXGRR2d7yNp1ykOaqNPzdVWn9roUyDVoG-Ai1_wktxC0_t_2PkWw8HB2iGpYBz2Bq2jwZ6y3v098A0tco8t
CitedBy_id crossref_primary_10_1177_14759217241243353
crossref_primary_10_1016_j_apt_2024_104743
crossref_primary_10_1109_ACCESS_2025_3566005
crossref_primary_10_3390_a17110489
crossref_primary_10_1016_j_measurement_2022_112302
crossref_primary_10_1016_j_aeue_2025_155791
crossref_primary_10_1049_tje2_70100
crossref_primary_10_1088_1361_6501_ace644
crossref_primary_10_1109_TPEL_2024_3459470
crossref_primary_10_3390_rs17152737
crossref_primary_10_1016_j_applthermaleng_2024_124693
crossref_primary_10_1016_j_engappai_2024_109400
crossref_primary_10_3390_app13042545
crossref_primary_10_1016_j_measurement_2025_117442
crossref_primary_10_1093_ijlct_ctae139
crossref_primary_10_3390_s22207835
crossref_primary_10_3724_j_issn_1007_2802_20240156
crossref_primary_10_1007_s40808_025_02353_8
crossref_primary_10_1016_j_aei_2022_101856
crossref_primary_10_1016_j_energy_2024_130326
crossref_primary_10_1016_j_imu_2023_101428
crossref_primary_10_3233_JIFS_230528
crossref_primary_10_32604_cmc_2025_060567
crossref_primary_10_1080_19475705_2022_2102942
crossref_primary_10_1016_j_agwat_2024_108972
crossref_primary_10_1016_j_jclepro_2022_134048
Cites_doi 10.4249/scholarpedia.5947
10.1007/s00371-019-01774-8
10.1038/323533a0
10.1162/neco.1997.9.8.1735
10.1109/5.726791
10.1109/CVPR.2017.113
10.1109/TNNLS.2016.2582924
10.1109/CVPR.2017.683
10.1117/1.JEI.27.3.033002
10.1109/TIP.2020.2965989
10.1007/s10617-019-09220-7
10.1109/CVPR.2015.7298594
10.1109/TGRS.2017.2755542
10.1162/neco.1989.1.4.541
10.1016/j.future.2017.09.073
10.1109/CVPR.2016.90
10.1007/s41095-020-0185-5
ContentType Journal Article
Copyright 2021 Elsevier Ltd
Copyright © 2021 Elsevier Ltd. All rights reserved.
Copyright_xml – notice: 2021 Elsevier Ltd
– notice: Copyright © 2021 Elsevier Ltd. All rights reserved.
DBID AAYXX
CITATION
7X8
DOI 10.1016/j.neunet.2021.08.032
DatabaseName CrossRef
MEDLINE - Academic
DatabaseTitle CrossRef
MEDLINE - Academic
DatabaseTitleList MEDLINE - Academic

Database_xml – sequence: 1
  dbid: 7X8
  name: MEDLINE - Academic
  url: https://search.proquest.com/medline
  sourceTypes: Aggregation Database
DeliveryMethod fulltext_linktorsrc
Discipline Computer Science
EISSN 1879-2782
EndPage 306
ExternalDocumentID 10_1016_j_neunet_2021_08_032
S0893608021003452
GroupedDBID ---
--K
--M
-~X
.DC
.~1
0R~
123
186
1B1
1RT
1~.
1~5
29N
4.4
457
4G.
53G
5RE
5VS
6TJ
7-5
71M
8P~
9JM
9JN
AABNK
AACTN
AADPK
AAEDT
AAEDW
AAIAV
AAIKJ
AAKOC
AALRI
AAOAW
AAQFI
AAQXK
AAXLA
AAXUO
AAYFN
ABAOU
ABBOA
ABCQJ
ABEFU
ABFNM
ABFRF
ABHFT
ABIVO
ABJNI
ABLJU
ABMAC
ABXDB
ABYKQ
ACAZW
ACDAQ
ACGFO
ACGFS
ACIUM
ACNNM
ACRLP
ACZNC
ADBBV
ADEZE
ADGUI
ADJOM
ADMUD
ADRHT
AEBSH
AECPX
AEFWE
AEKER
AENEX
AFKWA
AFTJW
AFXIZ
AGHFR
AGUBO
AGWIK
AGYEJ
AHHHB
AHJVU
AHZHX
AIALX
AIEXJ
AIKHN
AITUG
AJBFU
AJOXV
ALMA_UNASSIGNED_HOLDINGS
AMFUW
AMRAJ
AOUOD
ARUGR
ASPBG
AVWKF
AXJTR
AZFZN
BJAXD
BKOJK
BLXMC
CS3
DU5
EBS
EFJIC
EFLBG
EJD
EO8
EO9
EP2
EP3
F0J
F5P
FDB
FEDTE
FGOYB
FIRID
FNPLU
FYGXN
G-2
G-Q
G8K
GBLVA
GBOLZ
HLZ
HMQ
HVGLF
HZ~
IHE
J1W
JJJVA
K-O
KOM
KZ1
LG9
LMP
M2V
M41
MHUIS
MO0
MOBAO
MVM
N9A
O-L
O9-
OAUVE
OZT
P-8
P-9
P2P
PC.
Q38
R2-
RIG
ROL
RPZ
SBC
SCC
SDF
SDG
SDP
SES
SEW
SNS
SPC
SPCBC
SSN
SST
SSV
SSW
SSZ
T5K
TAE
UAP
UNMZH
VOH
WUQ
XPP
ZMT
~G-
9DU
AATTM
AAXKI
AAYWO
AAYXX
ABDPE
ABWVN
ACLOT
ACRPL
ACVFH
ADCNI
ADNMO
AEIPS
AEUPX
AFJKZ
AFPUW
AGQPQ
AIGII
AIIUN
AKBMS
AKRWK
AKYEP
ANKPU
APXCP
CITATION
EFKBS
~HD
7X8
ID FETCH-LOGICAL-c339t-154b00c496717f042102b8d86b1ff9019aede7c80965afd6440f57db2a351eae3
ISICitedReferencesCount 30
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=000704005300018&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN 0893-6080
1879-2782
IngestDate Mon Sep 29 02:49:40 EDT 2025
Tue Nov 18 21:08:39 EST 2025
Sat Nov 29 07:19:00 EST 2025
Fri Feb 23 02:41:14 EST 2024
IsPeerReviewed true
IsScholarly true
Keywords Deep learning
Sequence modeling
Parallel computation
Carry-lookahead
Language English
LinkModel OpenURL
MergedId FETCHMERGED-LOGICAL-c339t-154b00c496717f042102b8d86b1ff9019aede7c80965afd6440f57db2a351eae3
Notes ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 23
ORCID 0000-0003-1208-972X
0000-0002-5619-5252
0000-0002-9880-0342
0000-0002-3238-5430
PQID 2575068180
PQPubID 23479
PageCount 10
ParticipantIDs proquest_miscellaneous_2575068180
crossref_citationtrail_10_1016_j_neunet_2021_08_032
crossref_primary_10_1016_j_neunet_2021_08_032
elsevier_sciencedirect_doi_10_1016_j_neunet_2021_08_032
PublicationCentury 2000
PublicationDate December 2021
2021-12-00
20211201
PublicationDateYYYYMMDD 2021-12-01
PublicationDate_xml – month: 12
  year: 2021
  text: December 2021
PublicationDecade 2020
PublicationTitle Neural networks
PublicationYear 2021
Publisher Elsevier Ltd
Publisher_xml – name: Elsevier Ltd
References Zhang, Wu, Che, Lin, Memisevic, Salakhutdinov (b51) 2016
Graves (b10) 2013
Cho, Van Merriënboer, Gulcehre, Bahdanau, Bougares, Schwenk (b7) 2014
Pascanu, Gulcehre, Cho, Bengio (b37) 2013
Hochreiter, Schmidhuber (b14) 1997; 9
Bahdanau, Cho, Bengio (b2) 2014
Goodfellow, Bengio, Courville, Bengio (b9) 2016
Shazeer, Mirhoseini, Maziarz, Davis, Le, Hinton (b40) 2017
Krueger, Maharaj, Kramár, Pezeshki, Ballas, Ke (b23) 2016
Vohra, Goel, Sahoo (b45) 2015
Lea, C., Flynn, M. D., Vidal, R., Reiter, A., & Hager, G. D. (2017). Temporal convolutional networks for action segmentation and detection. In
Jing, Shen, Dubcek, Peurifoy, Skirlo, LeCun (b16) 2017
LeCun, Boser, Denker, Henderson, Howard, Hubbard (b29) 1989; 1
Rosenberger (b38) 1960
(pp. 156–165).
Sutskever, Vinyals, Le (b42) 2014
Jozefowicz, Zaremba, Sutskever (b18) 2015
Rumelhart, Hinton, Williams (b39) 1986; 323
LeCun, Bottou, Bengio, Haffner (b30) 1998; 86
Zhang, He, Luo, Xu, He (b50) 2020; 6
Bai, Kolter, Koltun (b3) 2018
Chang, Zhang, Han, Yu, Guo, Tan (b5) 2017; 30
Kirichenko, Vernik, Kamkar, Walter, Miller, Albu (b21) 2019; 29
Miyamoto, Cho (b35) 2016
Shi, Chen, Wang, Yeung, Wong, Woo (b41) 2015; 2015
He, K., Zhang, X., Ren, S., & Sun, J. (2016). Deep residual learning for image recognition. In
(pp. 3156–3164).
Greff, Srivastava, Koutník, Steunebrink, Schmidhuber (b11) 2016; 28
Krueger, Memisevic (b24) 2015
LeCun, Kanter, Solla (b31) 1991
Wu, Schuster, Chen, Le, Norouzi, Macherey (b48) 2016
Le, Jaitly, Hinton (b26) 2015
Kalchbrenner, Espeholt, Simonyan, Oord, Graves, Kavukcuoglu (b19) 2016
Jozefowicz, Vinyals, Schuster, Shazeer, Wu (b17) 2016
Allan, Williams (b1) 2005; 17
Zhong, Li, Luo, Chapman (b52) 2017; 56
Chung, Gulcehre, Cho, Bengio (b8) 2014
Le, Mikolov (b27) 2014
Zhang, He (b49) 2020; 36
Hinton (b13) 2009; 4
Mikolov, T., Sutskever, I., Deoras, A., Le, H.-S., Kombrink, S., & Cernocky, J. (2012). Subword language modeling with neural networks, preprint
(pp. 770–778).
Krizhevsky, Sutskever, Hinton (b22) 2012; 25
Zhou, He, Qiu (b54) 2018; 27
Chen, Tan, Wang, Lu, Hu, Fu (b6) 2020; 29
Wang, F., Jiang, M., Qian, C., Yang, S., Li, C., & Zhang, H., et al. (2017). Residual attention network for image classification. In
Szegedy, C., Liu, W., Jia, Y., Sermanet, P., Reed, S., & Anguelov, D., et al. (2015). Going deeper with convolutions. In
8, 67.
Kuchaiev, Ginsburg (b25) 2017
Oord, Dieleman, Zen, Simonyan, Vinyals, Graves (b36) 2016
(pp. 1–9).
Marcus, Santorini, Marcinkiewicz (b32) 1993; 19
Bradbury, Merity, Xiong, Socher (b4) 2016
Wisdom, Powers, Hershey, Roux, Atlas (b47) 2016
Thapliyal, Srinivas (b44) 2005
Merity, Keskar, Socher (b33) 2017
Zhou, He, Hou, Qiu (b53) 2018; 79
Hou, Yan, He (b15) 2019; 23
Kalchbrenner, Grefenstette, Blunsom (b20) 2014
Kuchaiev (10.1016/j.neunet.2021.08.032_b25) 2017
Krueger (10.1016/j.neunet.2021.08.032_b24) 2015
Bai (10.1016/j.neunet.2021.08.032_b3) 2018
10.1016/j.neunet.2021.08.032_b46
Greff (10.1016/j.neunet.2021.08.032_b11) 2016; 28
Jozefowicz (10.1016/j.neunet.2021.08.032_b18) 2015
Bahdanau (10.1016/j.neunet.2021.08.032_b2) 2014
Chen (10.1016/j.neunet.2021.08.032_b6) 2020; 29
10.1016/j.neunet.2021.08.032_b43
Wu (10.1016/j.neunet.2021.08.032_b48) 2016
Chung (10.1016/j.neunet.2021.08.032_b8) 2014
Sutskever (10.1016/j.neunet.2021.08.032_b42) 2014
Zhou (10.1016/j.neunet.2021.08.032_b54) 2018; 27
Krizhevsky (10.1016/j.neunet.2021.08.032_b22) 2012; 25
Pascanu (10.1016/j.neunet.2021.08.032_b37) 2013
Hinton (10.1016/j.neunet.2021.08.032_b13) 2009; 4
Krueger (10.1016/j.neunet.2021.08.032_b23) 2016
Zhang (10.1016/j.neunet.2021.08.032_b51) 2016
Goodfellow (10.1016/j.neunet.2021.08.032_b9) 2016
Cho (10.1016/j.neunet.2021.08.032_b7) 2014
10.1016/j.neunet.2021.08.032_b12
Hou (10.1016/j.neunet.2021.08.032_b15) 2019; 23
Shazeer (10.1016/j.neunet.2021.08.032_b40) 2017
Allan (10.1016/j.neunet.2021.08.032_b1) 2005; 17
Merity (10.1016/j.neunet.2021.08.032_b33) 2017
Zhang (10.1016/j.neunet.2021.08.032_b50) 2020; 6
LeCun (10.1016/j.neunet.2021.08.032_b29) 1989; 1
Zhou (10.1016/j.neunet.2021.08.032_b53) 2018; 79
Bradbury (10.1016/j.neunet.2021.08.032_b4) 2016
Rosenberger (10.1016/j.neunet.2021.08.032_b38) 1960
Chang (10.1016/j.neunet.2021.08.032_b5) 2017; 30
10.1016/j.neunet.2021.08.032_b28
Zhong (10.1016/j.neunet.2021.08.032_b52) 2017; 56
Kalchbrenner (10.1016/j.neunet.2021.08.032_b20) 2014
Marcus (10.1016/j.neunet.2021.08.032_b32) 1993; 19
Oord (10.1016/j.neunet.2021.08.032_b36) 2016
Le (10.1016/j.neunet.2021.08.032_b27) 2014
Rumelhart (10.1016/j.neunet.2021.08.032_b39) 1986; 323
Kirichenko (10.1016/j.neunet.2021.08.032_b21) 2019; 29
Thapliyal (10.1016/j.neunet.2021.08.032_b44) 2005
Jing (10.1016/j.neunet.2021.08.032_b16) 2017
Wisdom (10.1016/j.neunet.2021.08.032_b47) 2016
LeCun (10.1016/j.neunet.2021.08.032_b30) 1998; 86
LeCun (10.1016/j.neunet.2021.08.032_b31) 1991
Shi (10.1016/j.neunet.2021.08.032_b41) 2015; 2015
Graves (10.1016/j.neunet.2021.08.032_b10) 2013
Kalchbrenner (10.1016/j.neunet.2021.08.032_b19) 2016
Hochreiter (10.1016/j.neunet.2021.08.032_b14) 1997; 9
Le (10.1016/j.neunet.2021.08.032_b26) 2015
Miyamoto (10.1016/j.neunet.2021.08.032_b35) 2016
Zhang (10.1016/j.neunet.2021.08.032_b49) 2020; 36
10.1016/j.neunet.2021.08.032_b34
Jozefowicz (10.1016/j.neunet.2021.08.032_b17) 2016
Vohra (10.1016/j.neunet.2021.08.032_b45) 2015
References_xml – reference: He, K., Zhang, X., Ren, S., & Sun, J. (2016). Deep residual learning for image recognition. In
– year: 2014
  ident: b42
  article-title: Sequence to sequence learning with neural networks
– year: 2018
  ident: b3
  article-title: An empirical evaluation of generic convolutional and recurrent networks for sequence modeling
– volume: 17
  start-page: 25
  year: 2005
  end-page: 32
  ident: b1
  article-title: Harmonising chorales by probabilistic inference
  publication-title: Advances in Neural Information Processing Systems
– year: 2015
  ident: b26
  article-title: A simple way to initialize recurrent networks of rectified linear units
– year: 2016
  ident: b36
  article-title: WaveNet: A generative model for raw audio
– year: 2014
  ident: b8
  article-title: Empirical evaluation of gated recurrent neural networks on sequence modeling
– volume: 29
  start-page: 1
  year: 2019
  end-page: 7
  ident: b21
  article-title: ERSFQ 8-bit parallel arithmetic logic unit
  publication-title: IEEE Transactions on Applied Superconductivity
– start-page: 805
  year: 2005
  end-page: 817
  ident: b44
  article-title: A novel reversible TSG gate and its application for designing reversible carry look-ahead and other adder architectures
  publication-title: Asia-Pacific conference on advances in computer systems architecture
– year: 2013
  ident: b10
  article-title: Generating sequences with recurrent neural networks
– year: 2016
  ident: b35
  article-title: Gated word-character recurrent language model
– reference: Wang, F., Jiang, M., Qian, C., Yang, S., Li, C., & Zhang, H., et al. (2017). Residual attention network for image classification. In
– year: 2014
  ident: b2
  article-title: Neural machine translation by jointly learning to align and translate
– volume: 2015
  start-page: 802
  year: 2015
  end-page: 810
  ident: b41
  article-title: Convolutional LSTM network: A machine learning approach for precipitation nowcasting
  publication-title: Advances in Neural Information Processing Systems
– year: 1960
  ident: b38
  article-title: Simultaneous carry adder
– start-page: 1
  year: 2015
  end-page: 4
  ident: b45
  article-title: Modeling temporal dependencies in data using a DBN-LSTM
  publication-title: IEEE international conference on data science and advanced analytics
– volume: 9
  start-page: 1735
  year: 1997
  end-page: 1780
  ident: b14
  article-title: Long short-term memory
  publication-title: Neural Computation
– reference: (pp. 3156–3164).
– reference: Szegedy, C., Liu, W., Jia, Y., Sermanet, P., Reed, S., & Anguelov, D., et al. (2015). Going deeper with convolutions. In
– year: 2016
  ident: b51
  article-title: Architectural complexity measures of recurrent neural networks
– year: 2017
  ident: b33
  article-title: Regularizing and optimizing LSTM language models
– volume: 6
  start-page: 455
  year: 2020
  end-page: 466
  ident: b50
  article-title: Weight asynchronous update: Improving the diversity of filters in a deep convolutional network
  publication-title: Computational Visual Media
– reference: (pp. 156–165).
– start-page: 918
  year: 1991
  end-page: 924
  ident: b31
  article-title: Second order properties of error surfaces: Learning time and generalization
  publication-title: Advances in neural information processing systems, Vol. 3
– reference: ) 8, 67.
– start-page: 2342
  year: 2015
  end-page: 2350
  ident: b18
  article-title: An empirical exploration of recurrent network architectures
  publication-title: International conference on machine learning
– volume: 4
  start-page: 5947
  year: 2009
  ident: b13
  article-title: Deep belief networks
  publication-title: Scholarpedia
– volume: 56
  start-page: 847
  year: 2017
  end-page: 858
  ident: b52
  article-title: Spectral–spatial residual network for hyperspectral image classification: A 3-D deep learning framework
  publication-title: IEEE Transactions on Geoscience and Remote Sensing
– volume: 30
  start-page: 77
  year: 2017
  end-page: 87
  ident: b5
  article-title: Dilated recurrent neural networks
  publication-title: Advances in Neural Information Processing Systems
– year: 2016
  ident: b9
  article-title: Deep learning, Vol. 1
– volume: 28
  start-page: 2222
  year: 2016
  end-page: 2232
  ident: b11
  article-title: LSTM: A search space odyssey
  publication-title: IEEE Transactions on Neural Networks and Learning Systems
– volume: 23
  start-page: 57
  year: 2019
  end-page: 77
  ident: b15
  article-title: A survey on partitioning models, solution algorithms and algorithm parallelization for hardware/software co-design
  publication-title: Design Automation for Embedded System
– volume: 1
  start-page: 541
  year: 1989
  end-page: 551
  ident: b29
  article-title: Backpropagation applied to handwritten zip code recognition
  publication-title: Neural Computation
– year: 2016
  ident: b23
  article-title: Zoneout: Regularizing RNNs by randomly preserving hidden activations
– start-page: 1733
  year: 2017
  end-page: 1741
  ident: b16
  article-title: Tunable efficient unitary neural networks (EUNN) and their application to RNNs
  publication-title: International conference on machine learning
– volume: 25
  start-page: 1097
  year: 2012
  end-page: 1105
  ident: b22
  article-title: ImageNet classification with deep convolutional neural networks
  publication-title: Advances in Neural Information Processing Systems
– year: 2016
  ident: b4
  article-title: Quasi-recurrent neural networks
– year: 2013
  ident: b37
  article-title: How to construct deep recurrent neural networks
– reference: (pp. 1–9).
– volume: 79
  start-page: 473
  year: 2018
  end-page: 487
  ident: b53
  article-title: Parallel ant colony optimization on multi-core SIMD CPUs
  publication-title: Future Generation Computer Systems
– volume: 36
  start-page: 1797
  year: 2020
  end-page: 1808
  ident: b49
  article-title: DRCDN: learning deep residual convolutional dehazing networks
  publication-title: The Visual Computer
– reference: Lea, C., Flynn, M. D., Vidal, R., Reiter, A., & Hager, G. D. (2017). Temporal convolutional networks for action segmentation and detection. In
– volume: 86
  start-page: 2278
  year: 1998
  end-page: 2324
  ident: b30
  article-title: Gradient-based learning applied to document recognition
  publication-title: Proceedings of the IEEE
– reference: (pp. 770–778).
– volume: 19
  start-page: 313
  year: 1993
  end-page: 330
  ident: b32
  article-title: Building a large annotated corpus of English: The Penn Treebank
  publication-title: Computational Linguistics
– year: 2014
  ident: b20
  article-title: A convolutional neural network for modelling sentences
– year: 2016
  ident: b19
  article-title: Neural machine translation in linear time
– year: 2017
  ident: b40
  article-title: Outrageously large neural networks: The sparsely-gated mixture-of-experts layer
– year: 2014
  ident: b7
  article-title: Learning phrase representations using RNN encoder-decoder for statistical machine translation
– year: 2015
  ident: b24
  article-title: Regularizing RNNs by stabilizing activations
– volume: 323
  start-page: 533
  year: 1986
  end-page: 536
  ident: b39
  article-title: Learning representations by back-propagating errors
  publication-title: Nature
– volume: 27
  year: 2018
  ident: b54
  article-title: Accelerating image convolution filtering algorithms on integrated CPU–GPU architectures
  publication-title: Journal of Electronic Imaging
– reference: Mikolov, T., Sutskever, I., Deoras, A., Le, H.-S., Kombrink, S., & Cernocky, J. (2012). Subword language modeling with neural networks, preprint (
– start-page: 1188
  year: 2014
  end-page: 1196
  ident: b27
  article-title: Distributed representations of sentences and documents
  publication-title: International conference on machine learning
– year: 2017
  ident: b25
  article-title: Factorization tricks for LSTM networks
– year: 2016
  ident: b48
  article-title: Google’s neural machine translation system: Bridging the gap between human and machine translation
– volume: 29
  start-page: 3763
  year: 2020
  end-page: 3776
  ident: b6
  article-title: Reverse attention-based residual network for salient object detection
  publication-title: IEEE Transactions on Image Processing
– year: 2016
  ident: b17
  article-title: Exploring the limits of language modeling
– year: 2016
  ident: b47
  article-title: Full-capacity unitary recurrent neural networks
– year: 2016
  ident: 10.1016/j.neunet.2021.08.032_b4
– volume: 4
  start-page: 5947
  issue: 5
  year: 2009
  ident: 10.1016/j.neunet.2021.08.032_b13
  article-title: Deep belief networks
  publication-title: Scholarpedia
  doi: 10.4249/scholarpedia.5947
– volume: 36
  start-page: 1797
  issue: 9
  year: 2020
  ident: 10.1016/j.neunet.2021.08.032_b49
  article-title: DRCDN: learning deep residual convolutional dehazing networks
  publication-title: The Visual Computer
  doi: 10.1007/s00371-019-01774-8
– year: 2016
  ident: 10.1016/j.neunet.2021.08.032_b19
– year: 1960
  ident: 10.1016/j.neunet.2021.08.032_b38
– volume: 19
  start-page: 313
  issue: 2
  year: 1993
  ident: 10.1016/j.neunet.2021.08.032_b32
  article-title: Building a large annotated corpus of English: The Penn Treebank
  publication-title: Computational Linguistics
– volume: 323
  start-page: 533
  issue: 6088
  year: 1986
  ident: 10.1016/j.neunet.2021.08.032_b39
  article-title: Learning representations by back-propagating errors
  publication-title: Nature
  doi: 10.1038/323533a0
– year: 2014
  ident: 10.1016/j.neunet.2021.08.032_b7
– year: 2016
  ident: 10.1016/j.neunet.2021.08.032_b9
– year: 2016
  ident: 10.1016/j.neunet.2021.08.032_b47
– start-page: 2342
  year: 2015
  ident: 10.1016/j.neunet.2021.08.032_b18
  article-title: An empirical exploration of recurrent network architectures
– volume: 9
  start-page: 1735
  issue: 8
  year: 1997
  ident: 10.1016/j.neunet.2021.08.032_b14
  article-title: Long short-term memory
  publication-title: Neural Computation
  doi: 10.1162/neco.1997.9.8.1735
– volume: 86
  start-page: 2278
  issue: 11
  year: 1998
  ident: 10.1016/j.neunet.2021.08.032_b30
  article-title: Gradient-based learning applied to document recognition
  publication-title: Proceedings of the IEEE
  doi: 10.1109/5.726791
– year: 2016
  ident: 10.1016/j.neunet.2021.08.032_b51
– ident: 10.1016/j.neunet.2021.08.032_b28
  doi: 10.1109/CVPR.2017.113
– volume: 29
  start-page: 1
  issue: 5
  year: 2019
  ident: 10.1016/j.neunet.2021.08.032_b21
  article-title: ERSFQ 8-bit parallel arithmetic logic unit
  publication-title: IEEE Transactions on Applied Superconductivity
– year: 2013
  ident: 10.1016/j.neunet.2021.08.032_b37
– start-page: 805
  year: 2005
  ident: 10.1016/j.neunet.2021.08.032_b44
  article-title: A novel reversible TSG gate and its application for designing reversible carry look-ahead and other adder architectures
– volume: 28
  start-page: 2222
  issue: 10
  year: 2016
  ident: 10.1016/j.neunet.2021.08.032_b11
  article-title: LSTM: A search space odyssey
  publication-title: IEEE Transactions on Neural Networks and Learning Systems
  doi: 10.1109/TNNLS.2016.2582924
– year: 2016
  ident: 10.1016/j.neunet.2021.08.032_b17
– ident: 10.1016/j.neunet.2021.08.032_b46
  doi: 10.1109/CVPR.2017.683
– volume: 30
  start-page: 77
  year: 2017
  ident: 10.1016/j.neunet.2021.08.032_b5
  article-title: Dilated recurrent neural networks
  publication-title: Advances in Neural Information Processing Systems
– volume: 17
  start-page: 25
  year: 2005
  ident: 10.1016/j.neunet.2021.08.032_b1
  article-title: Harmonising chorales by probabilistic inference
  publication-title: Advances in Neural Information Processing Systems
– volume: 27
  issue: 3
  year: 2018
  ident: 10.1016/j.neunet.2021.08.032_b54
  article-title: Accelerating image convolution filtering algorithms on integrated CPU–GPU architectures
  publication-title: Journal of Electronic Imaging
  doi: 10.1117/1.JEI.27.3.033002
– year: 2014
  ident: 10.1016/j.neunet.2021.08.032_b8
– volume: 29
  start-page: 3763
  year: 2020
  ident: 10.1016/j.neunet.2021.08.032_b6
  article-title: Reverse attention-based residual network for salient object detection
  publication-title: IEEE Transactions on Image Processing
  doi: 10.1109/TIP.2020.2965989
– volume: 25
  start-page: 1097
  year: 2012
  ident: 10.1016/j.neunet.2021.08.032_b22
  article-title: ImageNet classification with deep convolutional neural networks
  publication-title: Advances in Neural Information Processing Systems
– volume: 23
  start-page: 57
  issue: 1
  year: 2019
  ident: 10.1016/j.neunet.2021.08.032_b15
  article-title: A survey on partitioning models, solution algorithms and algorithm parallelization for hardware/software co-design
  publication-title: Design Automation for Embedded System
  doi: 10.1007/s10617-019-09220-7
– year: 2017
  ident: 10.1016/j.neunet.2021.08.032_b25
– year: 2016
  ident: 10.1016/j.neunet.2021.08.032_b48
– year: 2016
  ident: 10.1016/j.neunet.2021.08.032_b36
– year: 2013
  ident: 10.1016/j.neunet.2021.08.032_b10
– year: 2014
  ident: 10.1016/j.neunet.2021.08.032_b2
– volume: 2015
  start-page: 802
  year: 2015
  ident: 10.1016/j.neunet.2021.08.032_b41
  article-title: Convolutional LSTM network: A machine learning approach for precipitation nowcasting
  publication-title: Advances in Neural Information Processing Systems
– start-page: 918
  year: 1991
  ident: 10.1016/j.neunet.2021.08.032_b31
  article-title: Second order properties of error surfaces: Learning time and generalization
– year: 2014
  ident: 10.1016/j.neunet.2021.08.032_b42
– start-page: 1
  year: 2015
  ident: 10.1016/j.neunet.2021.08.032_b45
  article-title: Modeling temporal dependencies in data using a DBN-LSTM
– ident: 10.1016/j.neunet.2021.08.032_b43
  doi: 10.1109/CVPR.2015.7298594
– year: 2018
  ident: 10.1016/j.neunet.2021.08.032_b3
– volume: 56
  start-page: 847
  issue: 2
  year: 2017
  ident: 10.1016/j.neunet.2021.08.032_b52
  article-title: Spectral–spatial residual network for hyperspectral image classification: A 3-D deep learning framework
  publication-title: IEEE Transactions on Geoscience and Remote Sensing
  doi: 10.1109/TGRS.2017.2755542
– volume: 1
  start-page: 541
  issue: 4
  year: 1989
  ident: 10.1016/j.neunet.2021.08.032_b29
  article-title: Backpropagation applied to handwritten zip code recognition
  publication-title: Neural Computation
  doi: 10.1162/neco.1989.1.4.541
– volume: 79
  start-page: 473
  year: 2018
  ident: 10.1016/j.neunet.2021.08.032_b53
  article-title: Parallel ant colony optimization on multi-core SIMD CPUs
  publication-title: Future Generation Computer Systems
  doi: 10.1016/j.future.2017.09.073
– year: 2017
  ident: 10.1016/j.neunet.2021.08.032_b40
– start-page: 1733
  year: 2017
  ident: 10.1016/j.neunet.2021.08.032_b16
  article-title: Tunable efficient unitary neural networks (EUNN) and their application to RNNs
– year: 2016
  ident: 10.1016/j.neunet.2021.08.032_b35
– start-page: 1188
  year: 2014
  ident: 10.1016/j.neunet.2021.08.032_b27
  article-title: Distributed representations of sentences and documents
– ident: 10.1016/j.neunet.2021.08.032_b34
– year: 2015
  ident: 10.1016/j.neunet.2021.08.032_b24
– ident: 10.1016/j.neunet.2021.08.032_b12
  doi: 10.1109/CVPR.2016.90
– volume: 6
  start-page: 455
  issue: 4
  year: 2020
  ident: 10.1016/j.neunet.2021.08.032_b50
  article-title: Weight asynchronous update: Improving the diversity of filters in a deep convolutional network
  publication-title: Computational Visual Media
  doi: 10.1007/s41095-020-0185-5
– year: 2016
  ident: 10.1016/j.neunet.2021.08.032_b23
– year: 2017
  ident: 10.1016/j.neunet.2021.08.032_b33
– year: 2014
  ident: 10.1016/j.neunet.2021.08.032_b20
– year: 2015
  ident: 10.1016/j.neunet.2021.08.032_b26
SSID ssj0006843
Score 2.5694387
Snippet The recurrent network architecture is a widely used model in sequence modeling, but its serial dependency hinders the computation parallelization, which makes...
SourceID proquest
crossref
elsevier
SourceType Aggregation Database
Enrichment Source
Index Database
Publisher
StartPage 297
SubjectTerms Carry-lookahead
Deep learning
Parallel computation
Sequence modeling
Title Recurrent neural network from adder’s perspective: Carry-lookahead RNN
URI https://dx.doi.org/10.1016/j.neunet.2021.08.032
https://www.proquest.com/docview/2575068180
Volume 144
WOSCitedRecordID wos000704005300018&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
hasFullText 1
inHoldings 1
isFullTextHit
isPrint
journalDatabaseRights – providerCode: PRVESC
  databaseName: ScienceDirect database
  customDbUrl:
  eissn: 1879-2782
  dateEnd: 99991231
  omitProxy: false
  ssIdentifier: ssj0006843
  issn: 0893-6080
  databaseCode: AIEXJ
  dateStart: 19950101
  isFulltext: true
  titleUrlDefault: https://www.sciencedirect.com
  providerName: Elsevier
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwtV1bb9MwFLa6jQdeYNzEGExGQvCAXK1JGtt7m6pWA1UBjQ6FJ8uJHSlTlZa2u_DvdxzbaXZBGw-8RJUVJ-05p58_H58LQh9qkh7xjITKeKviQhHGOOx5Cp5LBQyV1om0P8c0SVia8u-dzg-fC3M-pVXFLi_5_L-qGsZA2SZ19h_U3TwUBuAzKB2uoHa4Pkjxx8aDXtdcMrUqQQOVjfS2mSTSVgyhS1Ow2KdZGrfAQC4Wf8gUOLcEfFafj5OkTVyTa89axx2Wzt98JGcXumzcqLYywUiXrcGBtKc8ZbXGYzv518ytn879EPRaoRwWMRnlJKC2g1BX3zHmYTaK2kBpo3JvAbj1JZx2QULwi7rmjXWJVecEvVYvO_kmRifjsZgM08nH-W9iWomZI3fXV2UDbQW0zwHqtg6_DNOvzQIdM5t34b-lz6isw_5uv_hvjOXG2l0Tksk2euJ2EvjQWsAz1NHVc_TUd-nADrRfoFFjENgaBHZKxMYgcG0Qn5a4ZQ4H-IYxYDCGl-hkNJwMjojrnkHyMOQrAtwYIDWPeAw79gKwGahkxhSLs15RAAvkUitNc2bK_8hCAS_eL_pUZYEM-z0tdfgKbVazSr9GWPOY8T7tRSoEMmMaLgQ8y2mcZcE-p0rvoNBLSOSutLzpcDIVPobwVFi5CiNXYRqfhsEOIs2suS2tcs_91AtfOHpoaZ8A47ln5nuvKwHoaY7EZKVnZ0sRGCyKTb2DNw-4Zxc9Xv8F3qLN1eJMv0OP8vNVuVzsoQ2asj1nbFc5v5Bh
linkProvider Elsevier
openUrl ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=Recurrent+neural+network+from+adder%27s+perspective%3A+Carry-lookahead+RNN&rft.jtitle=Neural+networks&rft.au=Jiang%2C+Haowei&rft.au=Qin%2C+Feiwei&rft.au=Cao%2C+Jin&rft.au=Peng%2C+Yong&rft.date=2021-12-01&rft.issn=1879-2782&rft.eissn=1879-2782&rft.volume=144&rft.spage=297&rft_id=info:doi/10.1016%2Fj.neunet.2021.08.032&rft.externalDBID=NO_FULL_TEXT
thumbnail_l http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=0893-6080&client=summon
thumbnail_m http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=0893-6080&client=summon
thumbnail_s http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=0893-6080&client=summon