{"id":"https://openalex.org/W7157541216","doi":"https://doi.org/10.48550/arxiv.2604.24806","title":"Versioned Late Materialization for Ultra-Long Sequence Training in Recommendation Systems at Scale","display_name":"Versioned Late Materialization for Ultra-Long Sequence Training in Recommendation Systems at Scale","publication_year":2026,"publication_date":"2026-04-27","ids":{"openalex":"https://openalex.org/W7157541216","doi":"https://doi.org/10.48550/arxiv.2604.24806"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2604.24806","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.24806","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"article"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2604.24806","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5134822314","display_name":"Liang Guo","orcid":null},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Guo, Liang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134833709","display_name":"Ge Song","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Song, Ge","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134874414","display_name":"Litao Deng","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Deng, Litao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134844562","display_name":"Jianhui Sun","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Sun, Jianhui","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134867783","display_name":"Chufeng Hu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Hu, Chufeng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134850252","display_name":"Lu Zhang","orcid":"https://orcid.org/0000-0001-6035-5549"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhang, Lu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5027063174","display_name":"Zhen Ma","orcid":"https://orcid.org/0000-0001-6855-4622"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ma, Zhen","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134824459","display_name":"Shouwei Chen","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chen, Shouwei","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134828879","display_name":"Weiran Liu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liu, Weiran","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134826445","display_name":"Sarang Masti Sreeshylan","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Sreeshylan, Sarang Masti","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5134872427","display_name":"Xiaoxuan Meng","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Meng, Xiaoxuan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":11,"corresponding_author_ids":["https://openalex.org/A5134822314"],"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":0.1867000013589859,"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":0.1867000013589859,"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/T12026","display_name":"Explainable Artificial Intelligence (XAI)","score":0.1103999987244606,"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":0.07919999957084656,"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/redundancy","display_name":"Redundancy (engineering)","score":0.6470000147819519},{"id":"https://openalex.org/keywords/recommender-system","display_name":"Recommender system","score":0.5569999814033508},{"id":"https://openalex.org/keywords/data-redundancy","display_name":"Data redundancy","score":0.5411999821662903},{"id":"https://openalex.org/keywords/preprocessor","display_name":"Preprocessor","score":0.5264999866485596},{"id":"https://openalex.org/keywords/sequence","display_name":"Sequence (biology)","score":0.4740999937057495},{"id":"https://openalex.org/keywords/data-modeling","display_name":"Data modeling","score":0.4311999976634979},{"id":"https://openalex.org/keywords/throughput","display_name":"Throughput","score":0.4043000042438507},{"id":"https://openalex.org/keywords/training-set","display_name":"Training set","score":0.3991999924182892},{"id":"https://openalex.org/keywords/deep-learning","display_name":"Deep learning","score":0.34880000352859497}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7694000005722046},{"id":"https://openalex.org/C152124472","wikidata":"https://www.wikidata.org/wiki/Q1204361","display_name":"Redundancy (engineering)","level":2,"score":0.6470000147819519},{"id":"https://openalex.org/C557471498","wikidata":"https://www.wikidata.org/wiki/Q554950","display_name":"Recommender system","level":2,"score":0.5569999814033508},{"id":"https://openalex.org/C7545210","wikidata":"https://www.wikidata.org/wiki/Q838123","display_name":"Data redundancy","level":2,"score":0.5411999821662903},{"id":"https://openalex.org/C34736171","wikidata":"https://www.wikidata.org/wiki/Q918333","display_name":"Preprocessor","level":2,"score":0.5264999866485596},{"id":"https://openalex.org/C2778112365","wikidata":"https://www.wikidata.org/wiki/Q3511065","display_name":"Sequence (biology)","level":2,"score":0.4740999937057495},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.43639999628067017},{"id":"https://openalex.org/C67186912","wikidata":"https://www.wikidata.org/wiki/Q367664","display_name":"Data modeling","level":2,"score":0.4311999976634979},{"id":"https://openalex.org/C157764524","wikidata":"https://www.wikidata.org/wiki/Q1383412","display_name":"Throughput","level":3,"score":0.4043000042438507},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.3991999924182892},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3513000011444092},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.34880000352859497},{"id":"https://openalex.org/C2778755073","wikidata":"https://www.wikidata.org/wiki/Q10858537","display_name":"Scale (ratio)","level":2,"score":0.3215999901294708},{"id":"https://openalex.org/C10551718","wikidata":"https://www.wikidata.org/wiki/Q5227332","display_name":"Data pre-processing","level":2,"score":0.3206999897956848},{"id":"https://openalex.org/C2777211547","wikidata":"https://www.wikidata.org/wiki/Q17141490","display_name":"Training (meteorology)","level":2,"score":0.3199999928474426},{"id":"https://openalex.org/C57493831","wikidata":"https://www.wikidata.org/wiki/Q3134666","display_name":"Projection (relational algebra)","level":2,"score":0.3138999938964844},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.3025999963283539},{"id":"https://openalex.org/C99844830","wikidata":"https://www.wikidata.org/wiki/Q102441924","display_name":"Scaling","level":2,"score":0.3003999888896942},{"id":"https://openalex.org/C48044578","wikidata":"https://www.wikidata.org/wiki/Q727490","display_name":"Scalability","level":2,"score":0.29809999465942383},{"id":"https://openalex.org/C75684735","wikidata":"https://www.wikidata.org/wiki/Q858810","display_name":"Big data","level":2,"score":0.29589998722076416},{"id":"https://openalex.org/C79403827","wikidata":"https://www.wikidata.org/wiki/Q3988","display_name":"Real-time computing","level":1,"score":0.28619998693466187},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.28619998693466187},{"id":"https://openalex.org/C77088390","wikidata":"https://www.wikidata.org/wiki/Q8513","display_name":"Database","level":1,"score":0.27720001339912415},{"id":"https://openalex.org/C43214815","wikidata":"https://www.wikidata.org/wiki/Q7310987","display_name":"Reliability (semiconductor)","level":3,"score":0.2687000036239624},{"id":"https://openalex.org/C205711294","wikidata":"https://www.wikidata.org/wiki/Q176953","display_name":"Rendering (computer graphics)","level":2,"score":0.2671000063419342},{"id":"https://openalex.org/C82876162","wikidata":"https://www.wikidata.org/wiki/Q17096504","display_name":"Latency (audio)","level":2,"score":0.26440000534057617},{"id":"https://openalex.org/C206345919","wikidata":"https://www.wikidata.org/wiki/Q20380951","display_name":"Resource (disambiguation)","level":2,"score":0.2565000057220459},{"id":"https://openalex.org/C63479239","wikidata":"https://www.wikidata.org/wiki/Q7353546","display_name":"Robustness (evolution)","level":3,"score":0.2554999887943268}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2604.24806","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.24806","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"article"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2604.24806","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.24806","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"article"},"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/9","display_name":"Industry, innovation and infrastructure","score":0.6580913066864014}],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Modern":[0],"Deep":[1],"Learning":[2],"Recommendation":[3],"Models":[4],"(DLRMs)":[5],"follow":[6],"scaling":[7,175],"laws":[8],"with":[9,61,138],"sequence":[10,64,150,173],"length,":[11],"driving":[12],"the":[13,22,146,162,184],"frontier":[14],"toward":[15],"ultra-long":[16],"User":[17],"Interaction":[18],"History":[19],"(UIH).":[20],"However,":[21],"industry-standard":[23],"\"Fat":[24],"Row\"":[25],"paradigm,":[26],"which":[27],"pre-materializes":[28],"these":[29],"sequences":[30,93],"into":[31],"every":[32],"training":[33,47,96,153,165],"example,":[34],"creates":[35],"a":[36,68,73,87,108,122],"storage":[37,125],"and":[38,91,118,142,195],"I/O":[39,140],"wall":[40],"where":[41,59],"data":[42,51,136,166,186],"infrastructure":[43,167,187],"usage":[44,169],"exceeds":[45],"GPU":[46],"capacity":[48],"due":[49],"to":[50],"redundancy":[52,81],"that":[53,78,111,176],"is":[54],"amplified":[55],"in":[56,86],"multi-tenant":[57],"environments":[58],"models":[60],"vastly":[62],"different":[63],"length":[65,174],"requirements":[66],"share":[67],"union":[69],"dataset.":[70],"We":[71],"present":[72],"\\emph{versioned":[74],"late":[75],"materialization}":[76],"paradigm":[77],"eliminates":[79],"this":[80],"by":[82,156],"storing":[83],"UIH":[84],"once":[85],"normalized,":[88],"immutable":[89,124],"tier":[90],"reconstructing":[92],"just-in-time":[94],"during":[95],"via":[97],"lightweight":[98],"versioned":[99],"pointers.":[100],"The":[101],"system":[102,163],"ensures":[103],"Online-to-Offline":[104],"(O2O)":[105],"consistency":[106],"through":[107],"bifurcated":[109],"protocol":[110],"prevents":[112],"future":[113],"leakage":[114],"across":[115],"both":[116],"streaming":[117],"batch":[119],"training,":[120],"while":[121,170],"read-optimized":[123],"layer":[126],"provides":[127],"multi-dimensional":[128],"projection":[129],"pushdown":[130],"for":[131,188],"heterogeneous":[132],"model":[133,179,191],"tenants.":[134],"Disaggregated":[135],"preprocessing":[137],"pipelined":[139],"prefetching":[141],"data-affinity":[143],"optimizations":[144],"masks":[145],"latency":[147],"of":[148],"training-time":[149],"reconstruction,":[151],"keeping":[152],"throughput":[154],"compute-bound":[155],"GPUs.":[157],"Deployed":[158],"on":[159],"production":[160],"DLRMs,":[161],"reduces":[164],"resource":[168],"enabling":[171],"aggressive":[172],"delivers":[177],"significant":[178],"quality":[180],"gains,":[181],"serving":[182],"as":[183],"foundational":[185],"modern":[189],"recommendation":[190],"architectures,":[192],"including":[193],"HSTU":[194],"ULTRA-HSTU.":[196]},"counts_by_year":[],"updated_date":"2026-04-30T06:11:10.768123","created_date":"2026-04-30T00:00:00"}