{"id":"https://openalex.org/W2948201212","doi":"https://doi.org/10.18653/v1/w19-4819","title":"Hierarchical Representation in Neural Language Models: Suppression and Recovery of Expectations","display_name":"Hierarchical Representation in Neural Language Models: Suppression and Recovery of Expectations","publication_year":2019,"publication_date":"2019-01-01","ids":{"openalex":"https://openalex.org/W2948201212","doi":"https://doi.org/10.18653/v1/w19-4819","mag":"2948201212"},"language":"en","primary_location":{"id":"doi:10.18653/v1/w19-4819","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/w19-4819","pdf_url":"https://www.aclweb.org/anthology/W19-4819.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 2019 ACL Workshop BlackboxNLP: Analyzing and Interpreting Neural Networks for NLP","raw_type":"proceedings-article"},"type":"preprint","indexed_in":["arxiv","crossref","datacite"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://www.aclweb.org/anthology/W19-4819.pdf","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5011708753","display_name":"Ethan Wilcox","orcid":"https://orcid.org/0000-0002-5128-9890"},"institutions":[{"id":"https://openalex.org/I136199984","display_name":"Harvard University","ror":"https://ror.org/03vek6s52","country_code":"US","type":"education","lineage":["https://openalex.org/I136199984"]},{"id":"https://openalex.org/I2801851002","display_name":"Harvard University Press","ror":"https://ror.org/006v7bf86","country_code":"US","type":"other","lineage":["https://openalex.org/I136199984","https://openalex.org/I2801851002"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Ethan Wilcox","raw_affiliation_strings":["Department of Linguistics, Harvard University,","Harvard University ,"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Linguistics, Harvard University,","institution_ids":["https://openalex.org/I136199984"]},{"raw_affiliation_string":"Harvard University ,","institution_ids":["https://openalex.org/I2801851002"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5090215557","display_name":"Roger L\u00e9vy","orcid":"https://orcid.org/0000-0002-4493-8864"},"institutions":[{"id":"https://openalex.org/I4210104429","display_name":"Institute of Cognitive and Brain Sciences","ror":"https://ror.org/01c3w3270","country_code":"US","type":"education","lineage":["https://openalex.org/I4210104429"]},{"id":"https://openalex.org/I63966007","display_name":"Massachusetts Institute of Technology","ror":"https://ror.org/042nb2s44","country_code":"US","type":"education","lineage":["https://openalex.org/I63966007"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Roger Levy","raw_affiliation_strings":["Department of Brain and Cognitive Sciences, MIT,","Massachusetts Institute Of Technology#TAB#"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Brain and Cognitive Sciences, MIT,","institution_ids":["https://openalex.org/I4210104429"]},{"raw_affiliation_string":"Massachusetts Institute Of Technology#TAB#","institution_ids":["https://openalex.org/I63966007"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5050104206","display_name":"Richard Futrell","orcid":"https://orcid.org/0000-0003-4964-9524"},"institutions":[{"id":"https://openalex.org/I204250578","display_name":"University of California, Irvine","ror":"https://ror.org/04gyf1771","country_code":"US","type":"education","lineage":["https://openalex.org/I204250578"]},{"id":"https://openalex.org/I4210107233","display_name":"Language Science (South Korea)","ror":"https://ror.org/01h9v1373","country_code":"KR","type":"company","lineage":["https://openalex.org/I4210107233"]}],"countries":["KR","US"],"is_corresponding":false,"raw_author_name":"Richard Futrell","raw_affiliation_strings":["Department of Language Science, UC Irvine,","University of California, Irvine **"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Language Science, UC Irvine,","institution_ids":["https://openalex.org/I204250578","https://openalex.org/I4210107233"]},{"raw_affiliation_string":"University of California, Irvine **","institution_ids":["https://openalex.org/I204250578"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":3,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":0.2892,"has_fulltext":true,"cited_by_count":3,"citation_normalized_percentile":{"value":0.65634213,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":90,"max":96},"biblio":{"volume":null,"issue":null,"first_page":"181","last_page":"190"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","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/T10028","display_name":"Topic Modeling","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/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/T12031","display_name":"Speech and dialogue systems","score":0.9980999827384949,"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":0.7535693645477295},{"id":"https://openalex.org/keywords/context","display_name":"Context (archaeology)","score":0.6455370187759399},{"id":"https://openalex.org/keywords/natural-language","display_name":"Natural language","score":0.6363093852996826},{"id":"https://openalex.org/keywords/natural-language-processing","display_name":"Natural language processing","score":0.6050169467926025},{"id":"https://openalex.org/keywords/language-model","display_name":"Language model","score":0.5898870825767517},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.5764569640159607},{"id":"https://openalex.org/keywords/representation","display_name":"Representation (politics)","score":0.5455018281936646},{"id":"https://openalex.org/keywords/stack","display_name":"Stack (abstract data type)","score":0.5454996824264526},{"id":"https://openalex.org/keywords/embedding","display_name":"Embedding","score":0.47340402007102966},{"id":"https://openalex.org/keywords/language-identification","display_name":"Language identification","score":0.444308340549469},{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.4419747591018677},{"id":"https://openalex.org/keywords/suite","display_name":"Suite","score":0.439595103263855},{"id":"https://openalex.org/keywords/dependency","display_name":"Dependency (UML)","score":0.42542946338653564},{"id":"https://openalex.org/keywords/programming-language","display_name":"Programming language","score":0.10239598155021667}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7535693645477295},{"id":"https://openalex.org/C2779343474","wikidata":"https://www.wikidata.org/wiki/Q3109175","display_name":"Context (archaeology)","level":2,"score":0.6455370187759399},{"id":"https://openalex.org/C195324797","wikidata":"https://www.wikidata.org/wiki/Q33742","display_name":"Natural language","level":2,"score":0.6363093852996826},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.6050169467926025},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.5898870825767517},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5764569640159607},{"id":"https://openalex.org/C2776359362","wikidata":"https://www.wikidata.org/wiki/Q2145286","display_name":"Representation (politics)","level":3,"score":0.5455018281936646},{"id":"https://openalex.org/C9395851","wikidata":"https://www.wikidata.org/wiki/Q177929","display_name":"Stack (abstract data type)","level":2,"score":0.5454996824264526},{"id":"https://openalex.org/C41608201","wikidata":"https://www.wikidata.org/wiki/Q980509","display_name":"Embedding","level":2,"score":0.47340402007102966},{"id":"https://openalex.org/C129792486","wikidata":"https://www.wikidata.org/wiki/Q1050419","display_name":"Language identification","level":3,"score":0.444308340549469},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.4419747591018677},{"id":"https://openalex.org/C79581498","wikidata":"https://www.wikidata.org/wiki/Q1367530","display_name":"Suite","level":2,"score":0.439595103263855},{"id":"https://openalex.org/C19768560","wikidata":"https://www.wikidata.org/wiki/Q320727","display_name":"Dependency (UML)","level":2,"score":0.42542946338653564},{"id":"https://openalex.org/C199360897","wikidata":"https://www.wikidata.org/wiki/Q9143","display_name":"Programming language","level":1,"score":0.10239598155021667},{"id":"https://openalex.org/C17744445","wikidata":"https://www.wikidata.org/wiki/Q36442","display_name":"Political science","level":0,"score":0.0},{"id":"https://openalex.org/C166957645","wikidata":"https://www.wikidata.org/wiki/Q23498","display_name":"Archaeology","level":1,"score":0.0},{"id":"https://openalex.org/C86803240","wikidata":"https://www.wikidata.org/wiki/Q420","display_name":"Biology","level":0,"score":0.0},{"id":"https://openalex.org/C95457728","wikidata":"https://www.wikidata.org/wiki/Q309","display_name":"History","level":0,"score":0.0},{"id":"https://openalex.org/C199539241","wikidata":"https://www.wikidata.org/wiki/Q7748","display_name":"Law","level":1,"score":0.0},{"id":"https://openalex.org/C151730666","wikidata":"https://www.wikidata.org/wiki/Q7205","display_name":"Paleontology","level":1,"score":0.0},{"id":"https://openalex.org/C94625758","wikidata":"https://www.wikidata.org/wiki/Q7163","display_name":"Politics","level":2,"score":0.0}],"mesh":[],"locations_count":5,"locations":[{"id":"doi:10.18653/v1/w19-4819","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/w19-4819","pdf_url":"https://www.aclweb.org/anthology/W19-4819.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 2019 ACL Workshop BlackboxNLP: Analyzing and Interpreting Neural Networks for NLP","raw_type":"proceedings-article"},{"id":"pmh:oai:arXiv.org:1906.04068","is_oa":true,"landing_page_url":"http://arxiv.org/abs/1906.04068","pdf_url":"https://arxiv.org/pdf/1906.04068","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":"pmh:oai:dspace.mit.edu:1721.1/130418","is_oa":true,"landing_page_url":"https://hdl.handle.net/1721.1/130418","pdf_url":null,"source":{"id":"https://openalex.org/S4306400425","display_name":"DSpace@MIT (Massachusetts Institute of Technology)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I63966007","host_organization_name":"Massachusetts Institute of Technology","host_organization_lineage":["https://openalex.org/I63966007"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by-nc","license_id":"https://openalex.org/licenses/cc-by-nc","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"Association for Computational Linguistics","raw_type":"http://purl.org/eprint/type/ConferencePaper"},{"id":"doi:10.48550/arxiv.1906.04068","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.1906.04068","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":"article"},{"id":"mag:2948201212","is_oa":false,"landing_page_url":null,"pdf_url":null,"source":null,"license":null,"license_id":null,"version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":null}],"best_oa_location":{"id":"doi:10.18653/v1/w19-4819","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/w19-4819","pdf_url":"https://www.aclweb.org/anthology/W19-4819.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 2019 ACL Workshop BlackboxNLP: Analyzing and Interpreting Neural Networks for NLP","raw_type":"proceedings-article"},"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/4","score":0.8199999928474426,"display_name":"Quality Education"}],"awards":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W2948201212.pdf","grobid_xml":"https://content.openalex.org/works/W2948201212.grobid-xml"},"referenced_works_count":34,"referenced_works":["https://openalex.org/W1595256356","https://openalex.org/W1631260214","https://openalex.org/W2022741721","https://openalex.org/W2054125330","https://openalex.org/W2064675550","https://openalex.org/W2087946919","https://openalex.org/W2108010971","https://openalex.org/W2110485445","https://openalex.org/W2116876762","https://openalex.org/W2124479173","https://openalex.org/W2125001590","https://openalex.org/W2130035786","https://openalex.org/W2141845152","https://openalex.org/W2164418233","https://openalex.org/W2165131017","https://openalex.org/W2259472270","https://openalex.org/W2564486991","https://openalex.org/W2788924045","https://openalex.org/W2795342569","https://openalex.org/W2798579706","https://openalex.org/W2864832950","https://openalex.org/W2890027146","https://openalex.org/W2891399254","https://openalex.org/W2896556401","https://openalex.org/W2911109671","https://openalex.org/W2945185449","https://openalex.org/W2949908560","https://openalex.org/W2950191064","https://openalex.org/W2950311581","https://openalex.org/W2951714314","https://openalex.org/W2952817251","https://openalex.org/W2963073938","https://openalex.org/W2963751529","https://openalex.org/W2964222268"],"related_works":["https://openalex.org/W2972896975","https://openalex.org/W2943085977","https://openalex.org/W2956352737","https://openalex.org/W3058766270","https://openalex.org/W2948124660","https://openalex.org/W2952349142","https://openalex.org/W2963837967","https://openalex.org/W2952208026","https://openalex.org/W2340329281","https://openalex.org/W1978615613","https://openalex.org/W1573433453","https://openalex.org/W2798727047","https://openalex.org/W2250473257","https://openalex.org/W2741954161","https://openalex.org/W2251899728","https://openalex.org/W3019983303","https://openalex.org/W2485351821","https://openalex.org/W2984511769","https://openalex.org/W3016825020","https://openalex.org/W178964615"],"abstract_inverted_index":{"Deep":[0],"learning":[1],"sequence":[2],"models":[3,99,172],"have":[4],"led":[5],"to":[6,30,63,81,135,147],"a":[7,13,91,142,165],"marked":[8],"increase":[9],"in":[10,79,139,179],"performance":[11],"for":[12,35],"range":[14],"of":[15,56,85,93],"Natural":[16],"Language":[17],"Processing":[18],"tasks,":[19],"but":[20,182],"it":[21],"remains":[22],"an":[23],"open":[24],"question":[25],"whether":[26,96],"they":[27],"are":[28,54],"able":[29,146],"induce":[31,104],"proper":[32],"hierarchical":[33,83],"generalizations":[34],"representing":[36],"natural":[37,86,119],"language":[38,75,98,120],"from":[39],"linear":[40],"input":[41,49],"alone.":[42],"Work":[43],"using":[44],"artificial":[45],"languages":[46,70],"as":[47],"training":[48],"has":[50],"shown":[51],"that":[52,171],"LSTMs":[53],"capable":[55],"inducing":[57],"the":[58,82,130],"stack-like":[59,106],"data":[60,103,107],"structures":[61,84,108],"required":[62],"represent":[64],"context-free":[65],"and":[66,109,125,152,160,176],"certain":[67,150],"mildly":[68],"context-sensitive":[69],"(Weiss":[71],"et":[72],"al.,":[73],"2018)-formal":[74],"classes":[76],"which":[77],"correspond":[78],"theory":[80],"language.":[87],"Here":[88],"we":[89],"present":[90],"suite":[92],"experiments":[94],"probing":[95],"neural":[97],"trained":[100],"on":[101,129,164],"linguistic":[102],"these":[105,140,162],"deploy":[110],"them":[111],"while":[112],"incrementally":[113],"predicting":[114],"words.":[115],"We":[116],"study":[117],"two":[118],"phenomena:":[121],"center":[122],"embedding":[123],"sentences":[124],"syntactic":[126],"island":[127],"constraints":[128],"filler-gap":[131],"dependency.":[132],"In":[133],"order":[134],"properly":[136],"predict":[137],"words":[138],"structures,":[141],"model":[143],"must":[144],"be":[145],"temporarily":[148],"suppress":[149,175],"expectations":[151,156,163,178],"then":[153],"recover":[154,177,186],"those":[155],"later,":[157],"essentially":[158],"pushing":[159],"popping":[161],"stack.":[166],"Our":[167],"results":[168],"provide":[169],"evidence":[170],"can":[173],"successfully":[174],"many":[180],"cases,":[181],"do":[183],"not":[184],"fully":[185],"their":[187],"previous":[188],"grammatical":[189],"state.":[190]},"counts_by_year":[{"year":2020,"cited_by_count":2},{"year":2018,"cited_by_count":1}],"updated_date":"2026-06-11T09:08:48.828518","created_date":"2025-10-10T00:00:00"}