{"id":"https://openalex.org/W4382929605","doi":"https://doi.org/10.32657/10356/168487","title":"Natural language processing as autoregressive generation","display_name":"Natural language processing as autoregressive generation","publication_year":2023,"publication_date":"2023-01-01","ids":{"openalex":"https://openalex.org/W4382929605","doi":"https://doi.org/10.32657/10356/168487"},"language":"en","primary_location":{"id":"doi:10.32657/10356/168487","is_oa":true,"landing_page_url":"https://doi.org/10.32657/10356/168487","pdf_url":null,"source":null,"license":"cc-by-nc","license_id":"https://openalex.org/licenses/cc-by-nc","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Nanyang Technological University","raw_type":"dissertation"},"type":"dissertation","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://doi.org/10.32657/10356/168487","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5101792673","display_name":"Xiang Lin","orcid":"https://orcid.org/0000-0003-2789-4449"},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Lin, Xiang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":["https://openalex.org/A5101792673"],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"cited_by_count":1,"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":false,"primary_topic":{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.3495999872684479,"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":0.3495999872684479,"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":0.08910000324249268,"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/T14347","display_name":"Big Data and Digital Economy","score":0.05169999971985817,"subfield":{"id":"https://openalex.org/subfields/1710","display_name":"Information Systems"},"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/autoregressive-model","display_name":"Autoregressive model","score":0.6659353971481323},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.5510766506195068},{"id":"https://openalex.org/keywords/natural-language-processing","display_name":"Natural language processing","score":0.43592286109924316},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.35880425572395325},{"id":"https://openalex.org/keywords/econometrics","display_name":"Econometrics","score":0.15637370944023132},{"id":"https://openalex.org/keywords/mathematics","display_name":"Mathematics","score":0.13937273621559143}],"concepts":[{"id":"https://openalex.org/C159877910","wikidata":"https://www.wikidata.org/wiki/Q2202883","display_name":"Autoregressive model","level":2,"score":0.6659353971481323},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5510766506195068},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.43592286109924316},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.35880425572395325},{"id":"https://openalex.org/C149782125","wikidata":"https://www.wikidata.org/wiki/Q160039","display_name":"Econometrics","level":1,"score":0.15637370944023132},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.13937273621559143}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.32657/10356/168487","is_oa":true,"landing_page_url":"https://doi.org/10.32657/10356/168487","pdf_url":null,"source":null,"license":"cc-by-nc","license_id":"https://openalex.org/licenses/cc-by-nc","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Nanyang Technological University","raw_type":"dissertation"},{"id":"pmh:oai:dr.ntu.edu.sg:10356/168487","is_oa":true,"landing_page_url":"https://hdl.handle.net/10356/168487","pdf_url":null,"source":{"id":"https://openalex.org/S4306402609","display_name":"DR-NTU (Nanyang Technological University)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I172675005","host_organization_name":"Nanyang Technological University","host_organization_lineage":["https://openalex.org/I172675005"],"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":null,"raw_type":"Thesis-Doctor of Philosophy"}],"best_oa_location":{"id":"doi:10.32657/10356/168487","is_oa":true,"landing_page_url":"https://doi.org/10.32657/10356/168487","pdf_url":null,"source":null,"license":"cc-by-nc","license_id":"https://openalex.org/licenses/cc-by-nc","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Nanyang Technological University","raw_type":"dissertation"},"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/4","display_name":"Quality Education","score":0.6299999952316284}],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W4391375266","https://openalex.org/W2748952813","https://openalex.org/W2171218219","https://openalex.org/W1972271943","https://openalex.org/W2150410159","https://openalex.org/W2390279801","https://openalex.org/W2358668433","https://openalex.org/W4396701345","https://openalex.org/W4327525404","https://openalex.org/W3204019825"],"abstract_inverted_index":{"The":[0,155,192,312],"advances":[1],"in":[2,10,58,122,241,264,308,352,388],"deep":[3],"learning":[4,34,87],"have":[5,54,237],"led":[6],"to":[7,100,261,342,363,372,418,430,437],"great":[8,60],"achievements":[9],"many":[11,79],"Natural":[12],"Language":[13],"Processing":[14],"(NLP)":[15],"tasks.":[16,64,77,215,356],"With":[17],"the":[18,32,43,70,85,130,138,144,147,169,179,186,196,199,205,217,220,233,253,272,282,299,302,309,349,378,385,389,394,433,439,452],"nature":[19,197],"of":[20,42,62,90,140,146,153,185,198,252,294,301,404,454],"language,":[21],"i.e.,":[22,224],"sequential":[23],"data,":[24],"most":[25,44],"NLP":[26,49,63,76],"tasks":[27,80],"can":[28,81,360,450],"be":[29,361],"framed":[30],"into":[31,84,102,377],"sequence":[33,86],"framework,":[35,88],"such":[36],"as":[37,132],"text":[38,262,265,288,322,326,354,366,379],"generation.":[39,333],"As":[40],"one":[41,251],"important":[45],"foundations":[46],"for":[47,74,114,231,393],"modern":[48],"techniques,":[50],"autoregressive":[51,71,134,234],"generation":[52,72,135,235,323,355,380],"models":[53,236,456],"achieved":[55],"dominant":[56],"performance":[57,207],"a":[59,110,119,123,151,174,279,292,337,401,424],"deal":[61],"Therefore,":[65,105],"this":[66,106,398,443],"thesis":[67,107,399,444],"emphasizes":[68,109],"improving":[69],"model":[73,381],"different":[75,353,374],"While":[78],"naturally":[82],"fit":[83,101],"some":[89],"them,":[91],"e.g.,":[92],"building":[93],"discourse":[94,115,120,210],"parsing":[95,211,214],"tree,":[96],"require":[97],"sophisticated":[98],"designs":[99],"neural":[103,321],"models.":[104],"firstly":[108],"novel":[111,338,402,425],"unified":[112],"framework":[113,171],"parsing,":[116],"which":[117,177,249,297,346,411],"builds":[118],"tree":[121,200],"top-down":[124],"depth-first":[125],"manner,":[126],"and":[127,183,202,212,228,267,284,330,459],"it":[128],"frames":[129],"task":[131,136],"an":[133],"with":[137,161],"goal":[139],"each":[141],"step":[142],"being":[143],"prediction":[145],"node":[148],"position":[149],"given":[150],"piece":[152],"text.":[154],"proposed":[156,193,313,429],"approach":[157],"is":[158,250,428],"proven":[159],"effective":[160],"extensive":[162],"empirical":[163],"experiments.":[164],"In":[165,357],"addition,":[166],"I":[167,290,335],"extend":[168],"above":[170],"by":[172,305,382],"proposing":[173],"hierarchical":[175],"decoder,":[176],"leverages":[178],"information":[180,387],"from":[181,271],"parents":[182],"siblings":[184],"nodes":[187],"that":[188,446],"are":[189],"currently":[190],"processed.":[191],"decoder":[194],"utilizes":[195],"structure":[201],"further":[203,431],"improves":[204],"experiment":[206],"on":[208,319,408],"both":[209],"dependency":[213],"On":[216],"other":[218],"hand,":[219],"de":[221],"facto":[222],"strategies,":[223],"cross":[225,246,306,344],"entropy":[226,247],"loss":[227,427],"teacher":[229,268],"forcing,":[230],"training":[232,256,283,339,409],"been":[238],"shown":[239],"problematic":[240],"certain":[242],"aspects.":[243],"For":[244,287],"example,":[245],"loss,":[248],"widely":[254],"leveraged":[255],"objective":[257],"functions,":[258],"often":[259],"leads":[260],"degeneration":[263,350],"generation,":[266],"forcing":[269],"suffers":[270],"exposure":[273,395,448],"bias":[274,396,449],"problem,":[275,397],"where":[276],"there":[277],"exists":[278],"mismatch":[280],"between":[281],"testing":[285],"setup.":[286],"degeneration,":[289],"introduce":[291],"class":[293],"diminishing":[295,314],"attentions,":[296],"enforces":[298],"submodularity":[300],"coverage":[303],"calculated":[304],"attention":[307],"sequence-to-sequence":[310],"model.":[311],"attentions":[315],"achieve":[316],"notable":[317],"improvement":[318],"several":[320],"tasks,":[324],"including":[325],"summarization,":[327],"machine":[328],"translation,":[329],"image":[331],"paragraph":[332],"Further,":[334],"propose":[336],"objective,":[340],"ScaleGrad,":[341],"replace":[343],"entropy,":[345],"significantly":[347],"reduces":[348],"problem":[351],"fact,":[358],"ScaleGrad":[359],"extended":[362],"problems":[364],"beyond":[365],"degeneration.":[367],"It":[368],"provides":[369],"wide":[370],"flexibility":[371],"inject":[373],"inductive":[375],"biases":[376],"directly":[383],"modifying":[384],"gradient":[386],"output":[390],"layer.":[391],"Next,":[392],"introduces":[400],"type":[403],"scheduled":[405,420],"sampling":[406,421],"based":[407],"accuracy,":[410],"requires":[412],"only":[413],"minimal":[414],"hyper-parameter":[415],"tuning":[416],"compared":[417],"existing":[419],"methods.":[422],"Additionally,":[423],"imitation":[426],"enforce":[432],"model\u2019s":[434],"generative":[435],"behavior":[436],"match":[438],"teacher-forced":[440],"behavior.":[441],"Moreover,":[442],"demonstrates":[445],"reducing":[447],"improve":[451],"robustness":[453],"language":[455],"against":[457],"repetition":[458],"toxic":[460],"errors.":[461]},"counts_by_year":[{"year":2025,"cited_by_count":1}],"updated_date":"2026-03-25T13:04:00.132906","created_date":"2023-07-04T00:00:00"}