{"id":"https://openalex.org/W3087570533","doi":"https://doi.org/10.1109/tvcg.2020.3028976","title":"Attention Flows: Analyzing and Comparing Attention Mechanisms in Language Models","display_name":"Attention Flows: Analyzing and Comparing Attention Mechanisms in Language Models","publication_year":2020,"publication_date":"2020-10-14","ids":{"openalex":"https://openalex.org/W3087570533","doi":"https://doi.org/10.1109/tvcg.2020.3028976","mag":"3087570533","pmid":"https://pubmed.ncbi.nlm.nih.gov/33052855"},"language":"en","primary_location":{"id":"doi:10.1109/tvcg.2020.3028976","is_oa":false,"landing_page_url":"https://doi.org/10.1109/tvcg.2020.3028976","pdf_url":null,"source":{"id":"https://openalex.org/S84775595","display_name":"IEEE Transactions on Visualization and Computer Graphics","issn_l":"1077-2626","issn":["1077-2626","1941-0506","2160-9306"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319808","host_organization_name":"Institute of Electrical and Electronics Engineers","host_organization_lineage":["https://openalex.org/P4310319808"],"host_organization_lineage_names":["Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Transactions on Visualization and Computer Graphics","raw_type":"journal-article"},"type":"article","indexed_in":["arxiv","crossref","datacite","pubmed"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2009.07053","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5000073162","display_name":"Joseph F DeRose","orcid":null},"institutions":[{"id":"https://openalex.org/I200719446","display_name":"Vanderbilt University","ror":"https://ror.org/02vm5rt34","country_code":"US","type":"education","lineage":["https://openalex.org/I200719446"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Joseph F. DeRose","raw_affiliation_strings":["Vanderbilt University","Vanderbilt University,#TAB#"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Vanderbilt University","institution_ids":["https://openalex.org/I200719446"]},{"raw_affiliation_string":"Vanderbilt University,#TAB#","institution_ids":["https://openalex.org/I200719446"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101464031","display_name":"Jiayao Wang","orcid":"https://orcid.org/0000-0003-2953-9707"},"institutions":[{"id":"https://openalex.org/I200719446","display_name":"Vanderbilt University","ror":"https://ror.org/02vm5rt34","country_code":"US","type":"education","lineage":["https://openalex.org/I200719446"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Jiayao Wang","raw_affiliation_strings":["Vanderbilt University","Vanderbilt University,#TAB#"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Vanderbilt University","institution_ids":["https://openalex.org/I200719446"]},{"raw_affiliation_string":"Vanderbilt University,#TAB#","institution_ids":["https://openalex.org/I200719446"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5060051655","display_name":"Matthew Berger","orcid":"https://orcid.org/0000-0002-8876-2418"},"institutions":[{"id":"https://openalex.org/I200719446","display_name":"Vanderbilt University","ror":"https://ror.org/02vm5rt34","country_code":"US","type":"education","lineage":["https://openalex.org/I200719446"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Matthew Berger","raw_affiliation_strings":["Vanderbilt University","Vanderbilt University,#TAB#"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Vanderbilt University","institution_ids":["https://openalex.org/I200719446"]},{"raw_affiliation_string":"Vanderbilt University,#TAB#","institution_ids":["https://openalex.org/I200719446"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I200719446"],"apc_list":null,"apc_paid":null,"fwci":0.0964,"has_fulltext":true,"cited_by_count":2,"citation_normalized_percentile":{"value":0.4183507,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":89,"max":94},"biblio":{"volume":"27","issue":"2","first_page":"1160","last_page":"1170"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11714","display_name":"Multimodal Machine Learning Applications","score":0.9884999990463257,"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"}},"topics":[{"id":"https://openalex.org/T11714","display_name":"Multimodal Machine Learning Applications","score":0.9884999990463257,"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"}},{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.9778000116348267,"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/T12026","display_name":"Explainable Artificial Intelligence (XAI)","score":0.972100019454956,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.8209285736083984},{"id":"https://openalex.org/keywords/sentence","display_name":"Sentence","score":0.5652992129325867},{"id":"https://openalex.org/keywords/language-model","display_name":"Language model","score":0.5601175427436829},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.5236266255378723},{"id":"https://openalex.org/keywords/visual-analytics","display_name":"Visual analytics","score":0.4946117103099823},{"id":"https://openalex.org/keywords/natural-language-processing","display_name":"Natural language processing","score":0.48669248819351196},{"id":"https://openalex.org/keywords/variety","display_name":"Variety (cybernetics)","score":0.4787144660949707},{"id":"https://openalex.org/keywords/tracing","display_name":"Tracing","score":0.4637359380722046},{"id":"https://openalex.org/keywords/visualization","display_name":"Visualization","score":0.4547743499279022},{"id":"https://openalex.org/keywords/transformer","display_name":"Transformer","score":0.4360579252243042},{"id":"https://openalex.org/keywords/process","display_name":"Process (computing)","score":0.4266936480998993},{"id":"https://openalex.org/keywords/task","display_name":"Task (project management)","score":0.4232410490512848},{"id":"https://openalex.org/keywords/human\u2013computer-interaction","display_name":"Human\u2013computer interaction","score":0.34108424186706543},{"id":"https://openalex.org/keywords/machine-learning","display_name":"Machine learning","score":0.3311561048030853}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8209285736083984},{"id":"https://openalex.org/C2777530160","wikidata":"https://www.wikidata.org/wiki/Q41796","display_name":"Sentence","level":2,"score":0.5652992129325867},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.5601175427436829},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5236266255378723},{"id":"https://openalex.org/C59732488","wikidata":"https://www.wikidata.org/wiki/Q2528440","display_name":"Visual analytics","level":3,"score":0.4946117103099823},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.48669248819351196},{"id":"https://openalex.org/C136197465","wikidata":"https://www.wikidata.org/wiki/Q1729295","display_name":"Variety (cybernetics)","level":2,"score":0.4787144660949707},{"id":"https://openalex.org/C138673069","wikidata":"https://www.wikidata.org/wiki/Q322229","display_name":"Tracing","level":2,"score":0.4637359380722046},{"id":"https://openalex.org/C36464697","wikidata":"https://www.wikidata.org/wiki/Q451553","display_name":"Visualization","level":2,"score":0.4547743499279022},{"id":"https://openalex.org/C66322947","wikidata":"https://www.wikidata.org/wiki/Q11658","display_name":"Transformer","level":3,"score":0.4360579252243042},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.4266936480998993},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.4232410490512848},{"id":"https://openalex.org/C107457646","wikidata":"https://www.wikidata.org/wiki/Q207434","display_name":"Human\u2013computer interaction","level":1,"score":0.34108424186706543},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.3311561048030853},{"id":"https://openalex.org/C187736073","wikidata":"https://www.wikidata.org/wiki/Q2920921","display_name":"Management","level":1,"score":0.0},{"id":"https://openalex.org/C62520636","wikidata":"https://www.wikidata.org/wiki/Q944","display_name":"Quantum mechanics","level":1,"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/C165801399","wikidata":"https://www.wikidata.org/wiki/Q25428","display_name":"Voltage","level":2,"score":0.0},{"id":"https://openalex.org/C162324750","wikidata":"https://www.wikidata.org/wiki/Q8134","display_name":"Economics","level":0,"score":0.0},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","level":1,"score":0.0}],"mesh":[],"locations_count":5,"locations":[{"id":"doi:10.1109/tvcg.2020.3028976","is_oa":false,"landing_page_url":"https://doi.org/10.1109/tvcg.2020.3028976","pdf_url":null,"source":{"id":"https://openalex.org/S84775595","display_name":"IEEE Transactions on Visualization and Computer Graphics","issn_l":"1077-2626","issn":["1077-2626","1941-0506","2160-9306"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319808","host_organization_name":"Institute of Electrical and Electronics Engineers","host_organization_lineage":["https://openalex.org/P4310319808"],"host_organization_lineage_names":["Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Transactions on Visualization and Computer Graphics","raw_type":"journal-article"},{"id":"pmid:33052855","is_oa":false,"landing_page_url":"https://pubmed.ncbi.nlm.nih.gov/33052855","pdf_url":null,"source":{"id":"https://openalex.org/S4306525036","display_name":"PubMed","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I1299303238","host_organization_name":"National Institutes of Health","host_organization_lineage":["https://openalex.org/I1299303238"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE transactions on visualization and computer graphics","raw_type":null},{"id":"pmh:oai:arXiv.org:2009.07053","is_oa":true,"landing_page_url":"http://arxiv.org/abs/2009.07053","pdf_url":"https://arxiv.org/pdf/2009.07053","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"mag:3087570533","is_oa":true,"landing_page_url":"https://arxiv.org/abs/2009.07053v1","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"arXiv (Cornell University)","raw_type":null},{"id":"doi:10.48550/arxiv.2009.07053","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2009.07053","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2009.07053","is_oa":true,"landing_page_url":"http://arxiv.org/abs/2009.07053","pdf_url":"https://arxiv.org/pdf/2009.07053","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/4","display_name":"Quality Education","score":0.7200000286102295}],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":false},"content_urls":{"pdf":"https://content.openalex.org/works/W3087570533.pdf"},"referenced_works_count":54,"referenced_works":["https://openalex.org/W1840435438","https://openalex.org/W1968648267","https://openalex.org/W2052209137","https://openalex.org/W2064675550","https://openalex.org/W2157331557","https://openalex.org/W2160137030","https://openalex.org/W2194775991","https://openalex.org/W2250539671","https://openalex.org/W2343061342","https://openalex.org/W2740765036","https://openalex.org/W2751298778","https://openalex.org/W2751305043","https://openalex.org/W2752194699","https://openalex.org/W2889883176","https://openalex.org/W2923014074","https://openalex.org/W2923421605","https://openalex.org/W2942161347","https://openalex.org/W2947012480","https://openalex.org/W2950768109","https://openalex.org/W2950784811","https://openalex.org/W2951286828","https://openalex.org/W2951559648","https://openalex.org/W2953356739","https://openalex.org/W2962739339","https://openalex.org/W2963123635","https://openalex.org/W2963214037","https://openalex.org/W2963563735","https://openalex.org/W2969224142","https://openalex.org/W2969282568","https://openalex.org/W2970120757","https://openalex.org/W2971033911","https://openalex.org/W2972312591","https://openalex.org/W2972324944","https://openalex.org/W2973154008","https://openalex.org/W2996005046","https://openalex.org/W2999765337","https://openalex.org/W3011411500","https://openalex.org/W6607333740","https://openalex.org/W6631190155","https://openalex.org/W6680532216","https://openalex.org/W6682691769","https://openalex.org/W6727690538","https://openalex.org/W6739901393","https://openalex.org/W6745091552","https://openalex.org/W6745615791","https://openalex.org/W6755207826","https://openalex.org/W6766673545","https://openalex.org/W6767002648","https://openalex.org/W6767594909","https://openalex.org/W6768061436","https://openalex.org/W6768583413","https://openalex.org/W6773996589","https://openalex.org/W6780226713","https://openalex.org/W6948308983"],"related_works":["https://openalex.org/W3093452197","https://openalex.org/W3142802968","https://openalex.org/W2996005046","https://openalex.org/W3199088499","https://openalex.org/W3153758184","https://openalex.org/W3034740286","https://openalex.org/W3025341884","https://openalex.org/W2799142693","https://openalex.org/W2904901328","https://openalex.org/W3163142861","https://openalex.org/W2927333699","https://openalex.org/W2949362007","https://openalex.org/W2995942567","https://openalex.org/W2791656050","https://openalex.org/W3038035611","https://openalex.org/W2890353432","https://openalex.org/W3143979764","https://openalex.org/W2979949198","https://openalex.org/W3208983378","https://openalex.org/W2904706299"],"abstract_inverted_index":{"Advances":[0],"in":[1,89,102,116,153,189],"language":[2,22,27,91,118],"modeling":[3],"have":[4],"led":[5],"to":[6,52,86,99,185,199],"the":[7,54,141,154,159,168,201],"development":[8],"of":[9,20,57,161,203],"deep":[10],"attention-based":[11,90],"models":[12,28,175],"that":[13],"are":[14,29],"performant":[15],"across":[16,110],"a":[17,32,65,73,82,127,162],"wide":[18],"variety":[19],"natural":[21],"processing":[23],"(NLP)":[24],"problems.":[25],"These":[26],"typified":[30],"by":[31],"pre-training":[33],"process":[34],"on":[35,125,136],"large":[36],"unlabeled":[37],"text":[38],"corpora":[39],"and":[40,105,112,144,173,179,194],"subsequently":[41],"fine-tuned":[42,174],"for":[43,72],"specific":[44],"tasks.":[45,206],"Although":[46],"considerable":[47],"work":[48],"has":[49],"been":[50],"devoted":[51],"understanding":[53,87,192],"attention":[55,67,107,114,139,146,187,197],"mechanisms":[56,68,188],"pre-trained":[58,172],"models,":[59],"it":[60],"is":[61,97,130,134],"less":[62],"understood":[63],"how":[64,126,145,196],"model's":[66],"change":[69],"when":[70],"trained":[71],"target":[74],"NLP":[75],"task.":[76],"In":[77],"this":[78],"paper,":[79],"we":[80],"propose":[81],"visual":[83,169],"analytics":[84],"approach":[85],"fine-tuning":[88],"models.":[92,119],"Our":[93],"visualization,":[94],"Attention":[95,156,183],"Flows,":[96],"designed":[98],"support":[100],"users":[101,122],"querying,":[103],"tracing,":[104],"comparing":[106],"within":[108],"layers,":[109,111],"amongst":[113],"heads":[115],"Transformer-based":[117],"To":[120],"help":[121],"gain":[123],"insight":[124],"classification":[128],"decision":[129],"made,":[131],"our":[132],"design":[133],"centered":[135],"depicting":[137],"classification-based":[138],"at":[140],"deepest":[142],"layer":[143],"from":[147],"prior":[148],"layers":[149],"flows":[150],"throughout":[151],"words":[152],"input.":[155],"Flows":[157,184],"supports":[158],"analysis":[160],"single":[163],"model,":[164],"as":[165,167],"well":[166],"comparison":[170],"between":[171],"via":[176],"their":[177],"similarities":[178],"differences.":[180],"We":[181],"use":[182],"study":[186],"various":[190],"sentence":[191],"tasks":[193],"highlight":[195],"evolves":[198],"address":[200],"nuances":[202],"solving":[204],"these":[205]},"counts_by_year":[{"year":2024,"cited_by_count":1},{"year":2022,"cited_by_count":1}],"updated_date":"2026-07-15T18:14:33.161393","created_date":"2022-07-25T00:00:00"}