{"id":"https://openalex.org/W4416249449","doi":"https://doi.org/10.1109/ijcnn64981.2025.11228036","title":"Alleviating Directional Bias in Non-Autoregressive Transformers","display_name":"Alleviating Directional Bias in Non-Autoregressive Transformers","publication_year":2025,"publication_date":"2025-06-30","ids":{"openalex":"https://openalex.org/W4416249449","doi":"https://doi.org/10.1109/ijcnn64981.2025.11228036"},"language":null,"primary_location":{"id":"doi:10.1109/ijcnn64981.2025.11228036","is_oa":false,"landing_page_url":"https://doi.org/10.1109/ijcnn64981.2025.11228036","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 International Joint Conference on Neural Networks (IJCNN)","raw_type":"proceedings-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":false,"oa_status":"closed","oa_url":null,"any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5101654432","display_name":"Shuang Wang","orcid":"https://orcid.org/0000-0003-4321-9881"},"institutions":[{"id":"https://openalex.org/I204512498","display_name":"University of Macau","ror":"https://ror.org/01r4q9n85","country_code":"MO","type":"education","lineage":["https://openalex.org/I204512498"]}],"countries":["MO"],"is_corresponding":true,"raw_author_name":"Songsheng Wang","raw_affiliation_strings":["CT Lab, University of Macau,Macau,China"],"affiliations":[{"raw_affiliation_string":"CT Lab, University of Macau,Macau,China","institution_ids":["https://openalex.org/I204512498"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101542426","display_name":"Yu Wan","orcid":"https://orcid.org/0000-0001-9448-0671"},"institutions":[{"id":"https://openalex.org/I45928872","display_name":"Alibaba Group (China)","ror":"https://ror.org/00k642b80","country_code":"CN","type":"company","lineage":["https://openalex.org/I45928872"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yu Wan","raw_affiliation_strings":["Tongyi Lab, Alibaba Group Inc,Hangzhou,China"],"affiliations":[{"raw_affiliation_string":"Tongyi Lab, Alibaba Group Inc,Hangzhou,China","institution_ids":["https://openalex.org/I45928872"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101468579","display_name":"Derek F. Wong","orcid":"https://orcid.org/0000-0002-5307-7322"},"institutions":[{"id":"https://openalex.org/I204512498","display_name":"University of Macau","ror":"https://ror.org/01r4q9n85","country_code":"MO","type":"education","lineage":["https://openalex.org/I204512498"]}],"countries":["MO"],"is_corresponding":false,"raw_author_name":"Derek F. Wong","raw_affiliation_strings":["CT Lab, University of Macau,Macau,China"],"affiliations":[{"raw_affiliation_string":"CT Lab, University of Macau,Macau,China","institution_ids":["https://openalex.org/I204512498"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100669788","display_name":"Shijie Zhang","orcid":"https://orcid.org/0000-0002-9269-5406"},"institutions":[{"id":"https://openalex.org/I204512498","display_name":"University of Macau","ror":"https://ror.org/01r4q9n85","country_code":"MO","type":"education","lineage":["https://openalex.org/I204512498"]}],"countries":["MO"],"is_corresponding":false,"raw_author_name":"Shijie Zhang","raw_affiliation_strings":["CT Lab, University of Macau,Macau,China"],"affiliations":[{"raw_affiliation_string":"CT Lab, University of Macau,Macau,China","institution_ids":["https://openalex.org/I204512498"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101974545","display_name":"Yen\u2010Sheng Lin","orcid":"https://orcid.org/0000-0001-6283-8122"},"institutions":[{"id":"https://openalex.org/I4210165204","display_name":"Zhuhai Institute of Advanced Technology","ror":"https://ror.org/05r1mzq61","country_code":"CN","type":"facility","lineage":["https://openalex.org/I19820366","https://openalex.org/I4210145761","https://openalex.org/I4210165204"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yuchu Lin","raw_affiliation_strings":["DeepTranx,Zhuhai,China"],"affiliations":[{"raw_affiliation_string":"DeepTranx,Zhuhai,China","institution_ids":["https://openalex.org/I4210165204"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5025832925","display_name":"Lidia S. Chao","orcid":"https://orcid.org/0000-0001-6629-170X"},"institutions":[{"id":"https://openalex.org/I204512498","display_name":"University of Macau","ror":"https://ror.org/01r4q9n85","country_code":"MO","type":"education","lineage":["https://openalex.org/I204512498"]}],"countries":["MO"],"is_corresponding":false,"raw_author_name":"Lidia S. Chao","raw_affiliation_strings":["CT Lab, University of Macau,Macau,China"],"affiliations":[{"raw_affiliation_string":"CT Lab, University of Macau,Macau,China","institution_ids":["https://openalex.org/I204512498"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":6,"corresponding_author_ids":["https://openalex.org/A5101654432"],"corresponding_institution_ids":["https://openalex.org/I204512498"],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.18190779,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"1","last_page":"8"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10181","display_name":"Natural Language Processing Techniques","score":0.5091000199317932,"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":0.5091000199317932,"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":0.16189999878406525,"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/T10775","display_name":"Generative Adversarial Networks and Image Synthesis","score":0.08269999921321869,"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/transformer","display_name":"Transformer","score":0.7634999752044678},{"id":"https://openalex.org/keywords/security-token","display_name":"Security token","score":0.6378999948501587},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.6313999891281128},{"id":"https://openalex.org/keywords/machine-translation","display_name":"Machine translation","score":0.4650000035762787},{"id":"https://openalex.org/keywords/dependency","display_name":"Dependency (UML)","score":0.4383000135421753},{"id":"https://openalex.org/keywords/autoregressive-model","display_name":"Autoregressive model","score":0.40619999170303345},{"id":"https://openalex.org/keywords/dependency-graph","display_name":"Dependency graph","score":0.3127000033855438}],"concepts":[{"id":"https://openalex.org/C66322947","wikidata":"https://www.wikidata.org/wiki/Q11658","display_name":"Transformer","level":3,"score":0.7634999752044678},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6399000287055969},{"id":"https://openalex.org/C48145219","wikidata":"https://www.wikidata.org/wiki/Q1335365","display_name":"Security token","level":2,"score":0.6378999948501587},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.6313999891281128},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5246000289916992},{"id":"https://openalex.org/C203005215","wikidata":"https://www.wikidata.org/wiki/Q79798","display_name":"Machine translation","level":2,"score":0.4650000035762787},{"id":"https://openalex.org/C19768560","wikidata":"https://www.wikidata.org/wiki/Q320727","display_name":"Dependency (UML)","level":2,"score":0.4383000135421753},{"id":"https://openalex.org/C159877910","wikidata":"https://www.wikidata.org/wiki/Q2202883","display_name":"Autoregressive model","level":2,"score":0.40619999170303345},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.33719998598098755},{"id":"https://openalex.org/C16311509","wikidata":"https://www.wikidata.org/wiki/Q4148050","display_name":"Dependency graph","level":3,"score":0.3127000033855438},{"id":"https://openalex.org/C204030448","wikidata":"https://www.wikidata.org/wiki/Q101017","display_name":"Distillation","level":2,"score":0.31029999256134033},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.288100004196167},{"id":"https://openalex.org/C2779343474","wikidata":"https://www.wikidata.org/wiki/Q3109175","display_name":"Context (archaeology)","level":2,"score":0.28040000796318054},{"id":"https://openalex.org/C2779530757","wikidata":"https://www.wikidata.org/wiki/Q1207505","display_name":"Quality (philosophy)","level":2,"score":0.27000001072883606},{"id":"https://openalex.org/C155846161","wikidata":"https://www.wikidata.org/wiki/Q1143367","display_name":"Graphical model","level":2,"score":0.2678000032901764},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.2644999921321869},{"id":"https://openalex.org/C114289077","wikidata":"https://www.wikidata.org/wiki/Q3284399","display_name":"Statistical model","level":2,"score":0.25200000405311584}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/ijcnn64981.2025.11228036","is_oa":false,"landing_page_url":"https://doi.org/10.1109/ijcnn64981.2025.11228036","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 International Joint Conference on Neural Networks (IJCNN)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[{"id":"https://openalex.org/F4320321001","display_name":"National Natural Science Foundation of China","ror":"https://ror.org/01h0zpd94"},{"id":"https://openalex.org/F4320321655","display_name":"Science and Technology Development Fund","ror":"https://ror.org/044vr6g03"}],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":19,"referenced_works":["https://openalex.org/W2101105183","https://openalex.org/W2886776719","https://openalex.org/W2935811960","https://openalex.org/W2962915948","https://openalex.org/W2963736842","https://openalex.org/W2976965654","https://openalex.org/W2996987694","https://openalex.org/W3034999214","https://openalex.org/W3035490389","https://openalex.org/W3105214104","https://openalex.org/W3168577192","https://openalex.org/W3173691187","https://openalex.org/W3175665465","https://openalex.org/W3177172118","https://openalex.org/W3206689743","https://openalex.org/W4292474994","https://openalex.org/W4385245566","https://openalex.org/W4385572963","https://openalex.org/W4385696046"],"related_works":[],"abstract_inverted_index":{"Non-Autoregressive":[0],"Transformer":[1,47,106,119],"(NART)":[2],"has":[3],"emerged":[4],"as":[5],"a":[6,149,193],"promising":[7],"approach":[8],"for":[9],"fast":[10],"neural":[11],"machine":[12],"translation":[13,160,172],"due":[14],"to":[15,79,132,184,215],"its":[16],"independent":[17],"and":[18,83,103,116,158,201],"parallel":[19],"nature":[20],"during":[21],"inference.":[22],"Early":[23],"works":[24],"have":[25],"improved":[26],"NART":[27,38,197],"by":[28],"integrating":[29],"directional":[30,73],"token":[31,74,155],"dependency":[32,75,156],"information.":[33],"However,":[34],"the":[35,45,52,58,63,72,142,179,189,205],"outcome":[36],"of":[37,44,51,54,60,65,89,146,192,196,221],"is":[39,57],"still":[40],"unsatisfactory":[41],"against":[42,174],"those":[43],"Autoregressive":[46],"(ART)":[48],"models.":[49],"One":[50],"contributions":[53],"this":[55,91],"paper":[56,92],"observation":[59],"Directional":[61,134],"Bias,":[62],"propagation":[64],"exposure":[66],"bias":[67],"into":[68,97],"NAR":[69],"models":[70,126,176],"through":[71],"adopted,":[76],"which":[77,152],"leads":[78],"low":[80],"transition":[81],"quality":[82,173,191],"should":[84],"be":[85],"minimized.":[86],"In":[87],"light":[88],"this,":[90],"incorporates":[93],"future":[94],"context":[95],"information":[96],"both":[98],"Conventional":[99],"Knowledge":[100,113],"Distillation":[101,114],"(CKD)":[102],"Directed":[104],"Acyclic":[105],"(DA-T)":[107],"frameworks,":[108],"whereby":[109],"proposes":[110],"Bidirectional":[111,117],"Contextual":[112,118],"(BCKD)":[115],"(BC-T):":[120],"BCKD":[121,187,210],"employs":[122],"dual":[123],"AR":[124],"teacher":[125],"with":[127,148,209],"opposite":[128],"inference":[129],"directions":[130],"(L2R/R2L)":[131],"reduce":[133],"Bias":[135],"in":[136],"CKD":[137],"datasets,":[138],"while":[139,177],"BC-T":[140,169,206],"replaces":[141],"directed":[143],"acyclic":[144],"graph":[145],"DA-T":[147],"bidirectional":[150,154,162],"graph,":[151],"captures":[153],"information,":[157],"performs":[159],"via":[161],"ensemble":[163],"search.":[164],"Experimental":[165],"results":[166],"reveal":[167],"that":[168],"achieves":[170],"comparative":[171],"ART":[175,216],"preserving":[178],"high":[180],"generation":[181,190],"efficiency,":[182],"inherent":[183],"DA-T.":[185],"Furthermore,":[186],"enhances":[188],"diverse":[194],"spectrum":[195],"models,":[198,217],"including":[199],"GLAT":[200],"CMLM.":[202],"More":[203],"intriguingly,":[204],"model":[207],"equipped":[208],"exhibits":[211],"superior":[212],"performance":[213],"compared":[214],"achieving":[218],"an":[219],"improvement":[220],"0.97":[222],"BLEU":[223],"points.<sup":[224],"xmlns:mml=\"http://www.w3.org/1998/Math/MathML\"":[225],"xmlns:xlink=\"http://www.w3.org/1999/xlink\">1</sup>":[226]},"counts_by_year":[],"updated_date":"2026-04-09T08:11:56.329763","created_date":"2025-11-14T00:00:00"}