{"id":"https://openalex.org/W7163892141","doi":"https://doi.org/10.48550/arxiv.2606.07317","title":"Gated Bidirectional Linear Attention for Generative Retrieval","display_name":"Gated Bidirectional Linear Attention for Generative Retrieval","publication_year":2026,"publication_date":"2026-06-05","ids":{"openalex":"https://openalex.org/W7163892141","doi":"https://doi.org/10.48550/arxiv.2606.07317"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2606.07317","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.07317","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2606.07317","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5022052407","display_name":"Artem Matveev","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Matveev, Artem","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5119517679","display_name":"Vladislav Tytskiy","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Tytskiy, Vladislav","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5068390397","display_name":"Sergei Makeev","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Makeev, Sergei","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5119556708","display_name":"Sergei Liamaev","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liamaev, Sergei","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"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":false,"primary_topic":{"id":"https://openalex.org/T10203","display_name":"Recommender Systems and Techniques","score":0.31139999628067017,"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"}},"topics":[{"id":"https://openalex.org/T10203","display_name":"Recommender Systems and Techniques","score":0.31139999628067017,"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"}},{"id":"https://openalex.org/T10286","display_name":"Information Retrieval and Search Behavior","score":0.06909999996423721,"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"}},{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.05700000002980232,"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/encoder","display_name":"Encoder","score":0.6563000082969666},{"id":"https://openalex.org/keywords/key","display_name":"Key (lock)","score":0.46470001339912415},{"id":"https://openalex.org/keywords/bottleneck","display_name":"Bottleneck","score":0.4302999973297119},{"id":"https://openalex.org/keywords/focus","display_name":"Focus (optics)","score":0.424699991941452},{"id":"https://openalex.org/keywords/generative-grammar","display_name":"Generative grammar","score":0.3905999958515167},{"id":"https://openalex.org/keywords/mnist-database","display_name":"MNIST database","score":0.36329999566078186},{"id":"https://openalex.org/keywords/attention-network","display_name":"Attention network","score":0.361299991607666},{"id":"https://openalex.org/keywords/pooling","display_name":"Pooling","score":0.3391000032424927},{"id":"https://openalex.org/keywords/channel","display_name":"Channel (broadcasting)","score":0.3330000042915344}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8037999868392944},{"id":"https://openalex.org/C118505674","wikidata":"https://www.wikidata.org/wiki/Q42586063","display_name":"Encoder","level":2,"score":0.6563000082969666},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.46470001339912415},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4375},{"id":"https://openalex.org/C2780513914","wikidata":"https://www.wikidata.org/wiki/Q18210350","display_name":"Bottleneck","level":2,"score":0.4302999973297119},{"id":"https://openalex.org/C192209626","wikidata":"https://www.wikidata.org/wiki/Q190909","display_name":"Focus (optics)","level":2,"score":0.424699991941452},{"id":"https://openalex.org/C39890363","wikidata":"https://www.wikidata.org/wiki/Q36108","display_name":"Generative grammar","level":2,"score":0.3905999958515167},{"id":"https://openalex.org/C190502265","wikidata":"https://www.wikidata.org/wiki/Q17069496","display_name":"MNIST database","level":3,"score":0.36329999566078186},{"id":"https://openalex.org/C2993807640","wikidata":"https://www.wikidata.org/wiki/Q103709453","display_name":"Attention network","level":2,"score":0.361299991607666},{"id":"https://openalex.org/C70437156","wikidata":"https://www.wikidata.org/wiki/Q7228652","display_name":"Pooling","level":2,"score":0.3391000032424927},{"id":"https://openalex.org/C127162648","wikidata":"https://www.wikidata.org/wiki/Q16858953","display_name":"Channel (broadcasting)","level":2,"score":0.3330000042915344},{"id":"https://openalex.org/C68339613","wikidata":"https://www.wikidata.org/wiki/Q1549489","display_name":"Speedup","level":2,"score":0.328000009059906},{"id":"https://openalex.org/C159877910","wikidata":"https://www.wikidata.org/wiki/Q2202883","display_name":"Autoregressive model","level":2,"score":0.3271999955177307},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.326200008392334},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.3240000009536743},{"id":"https://openalex.org/C12713177","wikidata":"https://www.wikidata.org/wiki/Q1900281","display_name":"Perspective (graphical)","level":2,"score":0.3188999891281128},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.31139999628067017},{"id":"https://openalex.org/C34736171","wikidata":"https://www.wikidata.org/wiki/Q918333","display_name":"Preprocessor","level":2,"score":0.3084000051021576},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.29660001397132874},{"id":"https://openalex.org/C101738243","wikidata":"https://www.wikidata.org/wiki/Q786435","display_name":"Autoencoder","level":3,"score":0.2903999984264374},{"id":"https://openalex.org/C2779530757","wikidata":"https://www.wikidata.org/wiki/Q1207505","display_name":"Quality (philosophy)","level":2,"score":0.2896000146865845},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.28349998593330383},{"id":"https://openalex.org/C82876162","wikidata":"https://www.wikidata.org/wiki/Q17096504","display_name":"Latency (audio)","level":2,"score":0.2809999883174896},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.2797999978065491},{"id":"https://openalex.org/C167966045","wikidata":"https://www.wikidata.org/wiki/Q5532625","display_name":"Generative model","level":3,"score":0.27309998869895935},{"id":"https://openalex.org/C177148314","wikidata":"https://www.wikidata.org/wiki/Q170084","display_name":"Generalization","level":2,"score":0.2565999925136566},{"id":"https://openalex.org/C2777210771","wikidata":"https://www.wikidata.org/wiki/Q4927124","display_name":"Block (permutation group theory)","level":2,"score":0.2500999867916107}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2606.07317","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.07317","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2606.07317","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.07317","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"In":[0,25,55],"recommender":[1],"systems,":[2],"generative":[3],"retrieval":[4,179],"typically":[5],"uses":[6],"an":[7,10,18,150],"encoder-decoder":[8],"setup:":[9],"encoder":[11,41,63,120],"processes":[12],"a":[13,43,83,108,113,118,128,155],"user":[14],"interaction":[15],"history,":[16],"and":[17,107,125],"autoregressive":[19],"decoder":[20],"then":[21],"generates":[22],"recommended":[23],"items.":[24],"large-scale":[26,114],"streaming":[27],"services,":[28],"active":[29],"users":[30],"accumulate":[31],"very":[32],"long":[33],"histories":[34,38],"over":[35],"time.":[36],"As":[37],"grow,":[39],"the":[40,62,167],"becomes":[42],"major":[44],"latency":[45],"bottleneck":[46],"because":[47],"softmax":[48],"attention":[49,60,70,86,92],"scales":[50],"quadratically":[51],"with":[52,93],"sequence":[53],"length.":[54],"our":[56,173],"experiments,":[57],"using":[58],"bidirectional":[59,85,140],"in":[61,127],"substantially":[64],"improves":[65],"quality.":[66,142],"However,":[67],"most":[68],"sub-quadratic":[69],"methods":[71],"focus":[72],"on":[73,181],"causal":[74,98],"attention.":[75],"We":[76],"propose":[77],"Gated":[78],"Bidirectional":[79],"Linear":[80],"Attention":[81],"(GBLA),":[82],"linear-time":[84],"layer":[87],"that":[88,121,166],"extends":[89],"kernelized":[90],"linear":[91],"three":[94],"lightweight":[95],"components:":[96],"local":[97],"mixing":[99],"(Conv1D),":[100],"sequence-level":[101],"key":[102],"gating":[103],"for":[104],"soft":[105],"forgetting,":[106],"gated":[109],"RMSNorm":[110],"output.":[111],"On":[112,143],"Yandex":[115],"Music":[116],"dataset,":[117],"hybrid":[119,169],"interleaves":[122],"self-attention":[123,141,178],"(SA)":[124],"GBLA":[126,137,146],"1:2":[129],"ratio":[130],"(one":[131],"SA":[132],"block":[133],"followed":[134],"by":[135],"two":[136],"blocks)":[138],"matches":[139],"H100":[144],"GPUs,":[145],"reaches":[147],"up":[148],"to":[149,161],"$8.2\\times$":[151],"single-layer":[152],"speedup":[153],"at":[154],"history":[156],"length":[157],"of":[158],"32768,":[159],"compared":[160],"FlashAttention-v3.":[162],"Finally,":[163],"we":[164],"show":[165],"same":[168],"design":[170],"generalizes":[171],"beyond":[172],"proprietary":[174],"setting,":[175],"consistently":[176],"preserving":[177],"quality":[180],"public":[182],"Amazon":[183],"benchmarks.":[184]},"counts_by_year":[],"updated_date":"2026-07-28T07:46:37.118299","created_date":"2026-06-09T00:00:00"}