{"id":"https://openalex.org/W6922135576","doi":"https://doi.org/10.13016/dspace/6acj-ciwu","title":"Stronger Inductive Biases for Sample-Efficient and Controllable Neural Machine Translation","display_name":"Stronger Inductive Biases for Sample-Efficient and Controllable Neural Machine Translation","publication_year":2023,"publication_date":"2023-01-01","ids":{"openalex":"https://openalex.org/W6922135576","doi":"https://doi.org/10.13016/dspace/6acj-ciwu"},"language":"en","primary_location":{"id":"pmh:oai:drum.lib.umd.edu:1903/29910","is_oa":true,"landing_page_url":"http://hdl.handle.net/1903/29910","pdf_url":"https://drum.lib.umd.edu/bitstreams/1de20e1c-1f53-45b8-84c3-794f04f1690e/download","source":{"id":"https://openalex.org/S4306401518","display_name":"University Libraries (University of Maryland)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I66946132","host_organization_name":"University of Maryland, College Park","host_organization_lineage":["https://openalex.org/I66946132"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"","raw_type":"Dissertation"},"type":"other","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://drum.lib.umd.edu/bitstreams/1de20e1c-1f53-45b8-84c3-794f04f1690e/download","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":null,"display_name":"Xu, Weijia","orcid":null},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Xu, Weijia","raw_affiliation_strings":[],"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"cited_by_count":0,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":null,"last_page":null},"is_retracted":false,"is_paratext":false,"is_xpac":true,"primary_topic":null,"topics":[],"keywords":[{"id":"https://openalex.org/keywords/machine-translation","display_name":"Machine translation","score":0.7635999917984009},{"id":"https://openalex.org/keywords/spurious-relationship","display_name":"Spurious relationship","score":0.6424000263214111},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.5414999723434448},{"id":"https://openalex.org/keywords/component","display_name":"Component (thermodynamics)","score":0.5153999924659729},{"id":"https://openalex.org/keywords/control","display_name":"Control (management)","score":0.45159998536109924},{"id":"https://openalex.org/keywords/sequence","display_name":"Sequence (biology)","score":0.44609999656677246},{"id":"https://openalex.org/keywords/meaning","display_name":"Meaning (existential)","score":0.41260001063346863},{"id":"https://openalex.org/keywords/translation","display_name":"Translation (biology)","score":0.40709999203681946}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.767300009727478},{"id":"https://openalex.org/C203005215","wikidata":"https://www.wikidata.org/wiki/Q79798","display_name":"Machine translation","level":2,"score":0.7635999917984009},{"id":"https://openalex.org/C97256817","wikidata":"https://www.wikidata.org/wiki/Q1462316","display_name":"Spurious relationship","level":2,"score":0.6424000263214111},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.6227999925613403},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.5414999723434448},{"id":"https://openalex.org/C168167062","wikidata":"https://www.wikidata.org/wiki/Q1117970","display_name":"Component (thermodynamics)","level":2,"score":0.5153999924659729},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.49239999055862427},{"id":"https://openalex.org/C2775924081","wikidata":"https://www.wikidata.org/wiki/Q55608371","display_name":"Control (management)","level":2,"score":0.45159998536109924},{"id":"https://openalex.org/C2778112365","wikidata":"https://www.wikidata.org/wiki/Q3511065","display_name":"Sequence (biology)","level":2,"score":0.44609999656677246},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.4375},{"id":"https://openalex.org/C2780876879","wikidata":"https://www.wikidata.org/wiki/Q3054749","display_name":"Meaning (existential)","level":2,"score":0.41260001063346863},{"id":"https://openalex.org/C149364088","wikidata":"https://www.wikidata.org/wiki/Q185917","display_name":"Translation (biology)","level":4,"score":0.40709999203681946},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.37599998712539673},{"id":"https://openalex.org/C195324797","wikidata":"https://www.wikidata.org/wiki/Q33742","display_name":"Natural language","level":2,"score":0.35670000314712524},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.3517000079154968},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.3190000057220459},{"id":"https://openalex.org/C2779439875","wikidata":"https://www.wikidata.org/wiki/Q1078276","display_name":"Natural language understanding","level":3,"score":0.31839999556541443},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.3050999939441681},{"id":"https://openalex.org/C197352929","wikidata":"https://www.wikidata.org/wiki/Q1074074","display_name":"Inductive bias","level":4,"score":0.2694000005722046},{"id":"https://openalex.org/C2776187449","wikidata":"https://www.wikidata.org/wiki/Q1513879","display_name":"Natural language generation","level":3,"score":0.2628999948501587},{"id":"https://openalex.org/C198352243","wikidata":"https://www.wikidata.org/wiki/Q37105","display_name":"Line (geometry)","level":2,"score":0.26190000772476196},{"id":"https://openalex.org/C2780992000","wikidata":"https://www.wikidata.org/wiki/Q17016113","display_name":"Generator (circuit theory)","level":3,"score":0.25780001282691956},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.2549999952316284},{"id":"https://openalex.org/C2776608160","wikidata":"https://www.wikidata.org/wiki/Q4785462","display_name":"Natural (archaeology)","level":2,"score":0.25279998779296875}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:drum.lib.umd.edu:1903/29910","is_oa":true,"landing_page_url":"http://hdl.handle.net/1903/29910","pdf_url":"https://drum.lib.umd.edu/bitstreams/1de20e1c-1f53-45b8-84c3-794f04f1690e/download","source":{"id":"https://openalex.org/S4306401518","display_name":"University Libraries (University of Maryland)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I66946132","host_organization_name":"University of Maryland, College Park","host_organization_lineage":["https://openalex.org/I66946132"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"","raw_type":"Dissertation"},{"id":"doi:10.13016/dspace/6acj-ciwu","is_oa":true,"landing_page_url":"https://doi.org/10.13016/dspace/6acj-ciwu","pdf_url":null,"source":{"id":"https://openalex.org/S4306402644","display_name":"Digital Repository at the University of Maryland (University of Maryland College Park)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I66946132","host_organization_name":"University of Maryland, College Park","host_organization_lineage":["https://openalex.org/I66946132"],"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":"thesis"}],"best_oa_location":{"id":"pmh:oai:drum.lib.umd.edu:1903/29910","is_oa":true,"landing_page_url":"http://hdl.handle.net/1903/29910","pdf_url":"https://drum.lib.umd.edu/bitstreams/1de20e1c-1f53-45b8-84c3-794f04f1690e/download","source":{"id":"https://openalex.org/S4306401518","display_name":"University Libraries (University of Maryland)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I66946132","host_organization_name":"University of Maryland, College Park","host_organization_lineage":["https://openalex.org/I66946132"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"","raw_type":"Dissertation"},"sustainable_development_goals":[{"display_name":"Quality Education","score":0.8624446988105774,"id":"https://metadata.un.org/sdg/4"}],"awards":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W6922135576.pdf","grobid_xml":"https://content.openalex.org/works/W6922135576.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"As":[0],"one":[1],"of":[2,6,30,61,80,138,262,296,318,333,379,397],"the":[3,78,81,93,130,189,207,215,233,243,276,283,288,297,334,377,389,395],"oldest":[4],"applications":[5],"natural":[7],"language":[8,53,335],"processing,":[9],"machine":[10,44],"translation":[11,45,237,290],"(MT)":[12],"has":[13],"a":[14,27,183,194,224,293,303,343,349,453],"growing":[15],"impact":[16],"on":[17,51,58,352],"human":[18,62],"lives":[19],"both":[20,433],"as":[21,26,35],"an":[22,414],"end":[23,385],"application":[24],"and":[25,39,84,107,124,155,356,364,435,458],"key":[28],"component":[29],"cross-lingual":[31,36],"information":[32,37],"processing":[33],"such":[34],"retrieval":[38],"dialogue":[40],"generation.":[41],"Although":[42],"neural":[43],"(NMT)":[46],"models":[47,110,145,200,400],"achieve":[48],"impressive":[49],"performance":[50,368],"some":[52],"pairs,":[54],"they":[55,66,85],"are":[56,67,146,240,260],"trained":[57],"large":[59,294],"amounts":[60],"translations.":[63,159],"In":[64,134,314],"addition,":[65],"notorious":[68],"for":[69,89,330,456],"generating":[70],"fluent":[71],"outputs":[72,131,238,391,396],"that":[73,116,142,151,164,201,239,250,259,358,422],"do":[74],"not":[75],"faithfully":[76],"reflect":[77],"meaning":[79],"source":[82,244,298],"sentence,":[83],"make":[86],"it":[87,359],"difficult":[88,402],"users":[90,386,440],"to":[91,103,128,148,157,187,223,242,275,286,323,366,383,387,441],"control":[92,129,388],"outputs.":[94],"To":[95,246,438],"address":[96,188],"these":[97,168,272,311],"issues,":[98],"this":[99],"thesis":[100],"contributes":[101],"techniques":[102],"build":[104],"more":[105,392,446],"sample-efficient":[106,153],"controllable":[108],"NMT":[109,144,329,380,399,416],"by":[111,171,181,309],"incorporating":[112],"stronger":[113,173,325],"inductive":[114,174,326,374],"biases":[115,150,170,175,249,278,327,375],"help":[117],"correct":[118],"undesirable":[119,149,169],"biases,":[120],"integrate":[121,324],"prior":[122],"knowledge,":[123],"introduce":[125,342,452],"flexible":[126,447],"ways":[127],"in":[132,197,230,328,369,376,445,462],"NMT.":[133,463],"our":[135,315],"first":[136,255],"line":[137,317],"research,":[139,319],"we":[140,165,221,254,269,320,372,450],"show":[141,270,357],"current":[143],"susceptible":[147],"hinder":[152],"training":[154,177,185,216,280,408,434],"lead":[156],"unfaithful":[158],"We":[160,179,341,412],"further":[161,451],"provide":[162,302,442],"evidence":[163],"can":[166,360,423],"mitigate":[167],"integrating":[172,459],"through":[176],"algorithms.":[178],"start":[180],"introducing":[182],"new":[184],"objective":[186,347],"exposure":[190],"bias":[191],"problem":[192,196,229,235],"\u2014":[193,232,236],"common":[195],"sequence":[198,209],"generation":[199],"typically":[202],"causes":[203],"accumulated":[204],"errors":[205],"along":[206],"generated":[208],"at":[210,264,279,407,432],"inference":[211,265,410,436],"time,":[212,281],"especially":[213],"when":[214],"data":[217],"is":[218,401],"limited.":[219],"Next,":[220],"turn":[222],"well-known":[225],"but":[226],"less":[227],"studied":[228],"MT":[231],"hallucination":[234,252],"unrelated":[241],"text.":[245],"find":[247],"spurious":[248,277,312],"cause":[251],"errors,":[253],"identify":[256],"model":[257,284,381,390,417],"symptoms":[258,273],"indicative":[261],"hallucinations":[263,308],"time.":[266,411,437],"And":[267],"then,":[268],"how":[271,322],"connect":[274],"where":[282],"learns":[285],"predict":[287],"ground-truth":[289],"while":[291],"ignoring":[292],"part":[295],"sentence.":[299],"These":[300],"findings":[301],"future":[304],"path":[305],"toward":[306],"mitigating":[307],"addressing":[310],"biases.":[313],"second":[316],"study":[321,373],"effective":[331],"integration":[332],"priors":[336],"estimated":[337],"from":[338],"unsupervised":[339],"data.":[340],"novel":[344,419],"semi-supervised":[345],"learning":[346],"with":[348,403,418,428],"theoretical":[350],"guarantee":[351],"its":[353],"global":[354],"optimum":[355],"be":[361],"effectively":[362],"approximated":[363],"leads":[365],"improved":[367],"practice.":[370],"Finally,":[371],"form":[378],"architectures":[382],"allow":[384,439],"easily.":[393],"Controlling":[394],"standard":[398],"high":[404],"computational":[405,430],"cost":[406,431],"or":[409],"develop":[413],"edit-based":[415],"edit":[420],"operations":[421],"incorporate":[424],"users'":[425],"lexical":[426,443,460],"constraints":[427,444,461],"low":[429],"morphological":[448],"forms,":[449],"modular":[454],"framework":[455],"inflecting":[457]},"counts_by_year":[],"updated_date":"2026-03-20T23:20:44.827607","created_date":"2025-10-10T00:00:00"}