{"id":"https://openalex.org/W2988309730","doi":"https://doi.org/10.18653/v1/d19-5622","title":"Mixed Multi-Head Self-Attention for Neural Machine Translation","display_name":"Mixed Multi-Head Self-Attention for Neural Machine Translation","publication_year":2019,"publication_date":"2019-01-01","ids":{"openalex":"https://openalex.org/W2988309730","doi":"https://doi.org/10.18653/v1/d19-5622","mag":"2988309730"},"language":"en","primary_location":{"id":"doi:10.18653/v1/d19-5622","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/d19-5622","pdf_url":"https://www.aclweb.org/anthology/D19-5622.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 3rd Workshop on Neural Generation and Translation","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://www.aclweb.org/anthology/D19-5622.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5022128651","display_name":"Hongyi Cui","orcid":null},"institutions":[{"id":"https://openalex.org/I146399215","display_name":"University of Tsukuba","ror":"https://ror.org/02956yf07","country_code":"JP","type":"education","lineage":["https://openalex.org/I146399215"]}],"countries":["JP"],"is_corresponding":false,"raw_author_name":"Hongyi Cui","raw_affiliation_strings":["Graduate School of Systems and Information Engineering, University of Tsukuba, Japan"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Graduate School of Systems and Information Engineering, University of Tsukuba, Japan","institution_ids":["https://openalex.org/I146399215"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5062483495","display_name":"Shohei Iida","orcid":null},"institutions":[{"id":"https://openalex.org/I146399215","display_name":"University of Tsukuba","ror":"https://ror.org/02956yf07","country_code":"JP","type":"education","lineage":["https://openalex.org/I146399215"]}],"countries":["JP"],"is_corresponding":false,"raw_author_name":"Shohei Iida","raw_affiliation_strings":["Graduate School of Systems and Information Engineering, University of Tsukuba, Japan"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Graduate School of Systems and Information Engineering, University of Tsukuba, Japan","institution_ids":["https://openalex.org/I146399215"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5062582590","display_name":"Po-Hsuan Hung","orcid":null},"institutions":[{"id":"https://openalex.org/I146399215","display_name":"University of Tsukuba","ror":"https://ror.org/02956yf07","country_code":"JP","type":"education","lineage":["https://openalex.org/I146399215"]}],"countries":["JP"],"is_corresponding":false,"raw_author_name":"Po-Hsuan Hung","raw_affiliation_strings":["Graduate School of Systems and Information Engineering, University of Tsukuba, Japan"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Graduate School of Systems and Information Engineering, University of Tsukuba, Japan","institution_ids":["https://openalex.org/I146399215"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5066456246","display_name":"Takehito Utsuro","orcid":"https://orcid.org/0000-0003-4072-1833"},"institutions":[{"id":"https://openalex.org/I146399215","display_name":"University of Tsukuba","ror":"https://ror.org/02956yf07","country_code":"JP","type":"education","lineage":["https://openalex.org/I146399215"]}],"countries":["JP"],"is_corresponding":false,"raw_author_name":"Takehito Utsuro","raw_affiliation_strings":["Graduate School of Systems and Information Engineering, University of Tsukuba, Japan"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Graduate School of Systems and Information Engineering, University of Tsukuba, Japan","institution_ids":["https://openalex.org/I146399215"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5100520423","display_name":"Masaaki Nagata","orcid":null},"institutions":[{"id":"https://openalex.org/I2251713219","display_name":"NTT (Japan)","ror":"https://ror.org/00berct97","country_code":"JP","type":"company","lineage":["https://openalex.org/I2251713219"]}],"countries":["JP"],"is_corresponding":false,"raw_author_name":"Masaaki Nagata","raw_affiliation_strings":["NTT Communication Science Laboratories, NTT Corporation, Japan"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"NTT Communication Science Laboratories, NTT Corporation, Japan","institution_ids":["https://openalex.org/I2251713219"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":1.2028,"has_fulltext":true,"cited_by_count":15,"citation_normalized_percentile":{"value":0.85214313,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":91,"max":98},"biblio":{"volume":null,"issue":null,"first_page":"206","last_page":"214"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10181","display_name":"Natural Language Processing Techniques","score":0.9998999834060669,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T10181","display_name":"Natural Language Processing Techniques","score":0.9998999834060669,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.9998000264167786,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T11714","display_name":"Multimodal Machine Learning Applications","score":0.9901999831199646,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/machine-translation","display_name":"Machine translation","score":0.8361880779266357},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.8202531337738037},{"id":"https://openalex.org/keywords/transformer","display_name":"Transformer","score":0.7270063161849976},{"id":"https://openalex.org/keywords/redundancy","display_name":"Redundancy (engineering)","score":0.650810182094574},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.5940276980400085},{"id":"https://openalex.org/keywords/natural-language-processing","display_name":"Natural language processing","score":0.46317723393440247},{"id":"https://openalex.org/keywords/machine-learning","display_name":"Machine learning","score":0.4283071756362915}],"concepts":[{"id":"https://openalex.org/C203005215","wikidata":"https://www.wikidata.org/wiki/Q79798","display_name":"Machine translation","level":2,"score":0.8361880779266357},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8202531337738037},{"id":"https://openalex.org/C66322947","wikidata":"https://www.wikidata.org/wiki/Q11658","display_name":"Transformer","level":3,"score":0.7270063161849976},{"id":"https://openalex.org/C152124472","wikidata":"https://www.wikidata.org/wiki/Q1204361","display_name":"Redundancy (engineering)","level":2,"score":0.650810182094574},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5940276980400085},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.46317723393440247},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.4283071756362915},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","level":1,"score":0.0},{"id":"https://openalex.org/C165801399","wikidata":"https://www.wikidata.org/wiki/Q25428","display_name":"Voltage","level":2,"score":0.0},{"id":"https://openalex.org/C121332964","wikidata":"https://www.wikidata.org/wiki/Q413","display_name":"Physics","level":0,"score":0.0},{"id":"https://openalex.org/C62520636","wikidata":"https://www.wikidata.org/wiki/Q944","display_name":"Quantum mechanics","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.18653/v1/d19-5622","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/d19-5622","pdf_url":"https://www.aclweb.org/anthology/D19-5622.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 3rd Workshop on Neural Generation and Translation","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.18653/v1/d19-5622","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/d19-5622","pdf_url":"https://www.aclweb.org/anthology/D19-5622.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 3rd Workshop on Neural Generation and Translation","raw_type":"proceedings-article"},"sustainable_development_goals":[{"score":0.5,"display_name":"Sustainable cities and communities","id":"https://metadata.un.org/sdg/11"}],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W2988309730.pdf","grobid_xml":"https://content.openalex.org/works/W2988309730.grobid-xml"},"referenced_works_count":46,"referenced_works":["https://openalex.org/W222053410","https://openalex.org/W1522301498","https://openalex.org/W1902237438","https://openalex.org/W2101105183","https://openalex.org/W2130942839","https://openalex.org/W2131774270","https://openalex.org/W2133564696","https://openalex.org/W2157331557","https://openalex.org/W2183341477","https://openalex.org/W2194775991","https://openalex.org/W2613904329","https://openalex.org/W2759088880","https://openalex.org/W2859444450","https://openalex.org/W2866343820","https://openalex.org/W2886845974","https://openalex.org/W2888539709","https://openalex.org/W2908336025","https://openalex.org/W2912351236","https://openalex.org/W2933138175","https://openalex.org/W2940744433","https://openalex.org/W2946794439","https://openalex.org/W2949973181","https://openalex.org/W2950023669","https://openalex.org/W2951563833","https://openalex.org/W2951977278","https://openalex.org/W2952576443","https://openalex.org/W2952809536","https://openalex.org/W2962784628","https://openalex.org/W2962788148","https://openalex.org/W2962822108","https://openalex.org/W2963403868","https://openalex.org/W2963919854","https://openalex.org/W2963925437","https://openalex.org/W2964045208","https://openalex.org/W2964121744","https://openalex.org/W2964189376","https://openalex.org/W2964265128","https://openalex.org/W2964302946","https://openalex.org/W2964308564","https://openalex.org/W2970565456","https://openalex.org/W2977587102","https://openalex.org/W3037932933","https://openalex.org/W3041866211","https://openalex.org/W3204130541","https://openalex.org/W4323654151","https://openalex.org/W4385245566"],"related_works":["https://openalex.org/W3176018525","https://openalex.org/W2775554247","https://openalex.org/W3026554633","https://openalex.org/W2903810591","https://openalex.org/W2888520903","https://openalex.org/W2963499882","https://openalex.org/W3098873988","https://openalex.org/W2890256614","https://openalex.org/W2949454572","https://openalex.org/W2970247882"],"abstract_inverted_index":{"Recently,":[0],"the":[1,8,22,30,73,92,125,154,171],"Transformer":[2,173],"becomes":[3],"a":[4,99],"stateof-the-art":[5],"architecture":[6],"in":[7,58,63,77,119],"filed":[9],"of":[10,18,94,156],"neural":[11],"machine":[12],"translation":[13,147],"(NMT).":[14],"A":[15],"key":[16],"point":[17],"its":[19],"high-performance":[20],"is":[21,26,43],"multi-head":[23,81,95],"selfattention":[24],"which":[25,105],"supposed":[27],"to":[28,32,35,47,90,127],"allow":[29],"model":[31,126],"independently":[33],"attend":[34],"information":[36],"from":[37],"different":[38,50,55,120],"representation":[39],"subspaces.":[40],"However,":[41],"there":[42],"no":[44],"explicit":[45],"mechanism":[46],"ensure":[48],"that":[49,71],"attention":[51,76,112,118,121],"heads":[52,79],"indeed":[53],"capture":[54],"features,":[56],"and":[57,110,116,162],"practice,":[59],"redundancy":[60],"has":[61],"occurred":[62],"multiple":[64,78,133],"heads.":[65,122,134],"In":[66,88,135],"this":[67],"paper,":[68],"we":[69,97,149],"argue":[70],"using":[72],"same":[74],"global":[75,109],"limits":[80],"self-attention's":[82],"capacity":[83],"for":[84],"learning":[85],"distinct":[86,129],"features.":[87],"order":[89],"improve":[91],"expressiveness":[93],"self-attention,":[96],"propose":[98],"novel":[100],"Mixed":[101],"Multi-Head":[102],"Self-Attention":[103],"(MMA)":[104],"models":[106,159],"not":[107],"only":[108],"local":[111],"but":[113],"also":[114],"forward":[115],"backward":[117],"This":[123],"enables":[124],"learn":[128],"representations":[130],"explicitly":[131],"among":[132],"our":[136,158],"experiments":[137],"on":[138,168],"both":[139],"WAT17":[140],"English-Japanese":[141],"as":[142,144],"well":[143],"IWSLT14":[145],"German-English":[146],"task,":[148],"show":[150],"that,":[151],"without":[152],"increasing":[153],"number":[155],"parameters,":[157],"yield":[160],"consistent":[161],"significant":[163],"improvements":[164],"(0.9":[165],"BLEU":[166],"scores":[167],"average)":[169],"over":[170],"strong":[172],"baseline.":[174],"1":[175]},"counts_by_year":[{"year":2025,"cited_by_count":1},{"year":2023,"cited_by_count":3},{"year":2022,"cited_by_count":2},{"year":2021,"cited_by_count":5},{"year":2020,"cited_by_count":4}],"updated_date":"2026-07-29T14:22:42.915294","created_date":"2025-10-10T00:00:00"}