{"id":"https://openalex.org/W7047635990","doi":null,"title":"Granularized Weights: Matryoshka Representations for Nested Transformer Blocks","display_name":"Granularized Weights: Matryoshka Representations for Nested Transformer Blocks","publication_year":2025,"publication_date":"2025-01-01","ids":{"openalex":"https://openalex.org/W7047635990"},"language":"en","primary_location":{"id":"pmh:oai:DiVA.org:kth-365780","doi":null,"is_oa":true,"landing_page_url":"http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-365780","pdf_url":"https://kth.diva-portal.org/smash/get/diva2:1979081/FULLTEXT01","source":{"id":"https://openalex.org/S4306401559","display_name":"KTH Publication Database DiVA (KTH Royal Institute of Technology)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I86987016","host_organization_name":"KTH Royal Institute of Technology","host_organization_lineage":["https://openalex.org/I86987016"],"host_organization_lineage_names":[],"type":"repository"},"license":"other-oa","license_id":"https://openalex.org/licenses/other-oa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"info:eu-repo/semantics/bachelorThesis"},"volume":null,"type":"dissertation","is_traditional":true,"indexed_in":[],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://kth.diva-portal.org/smash/get/diva2:1979081/FULLTEXT01","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5132737880","display_name":"Axel Riley","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Riley, Axel","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5132738940","display_name":"Edvard Oxelstr\u00f6m","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Oxelstr\u00f6m, Edvard","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":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":{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","score":0.9878000020980835,"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"}},"topics":[{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","score":0.9878000020980835,"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"}},{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.006500000134110451,"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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.002199999988079071,"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/model-compression","display_name":"model compression","score":0.6000000238418579},{"id":"https://openalex.org/keywords/parameter-reduction","display_name":"parameter reduction","score":0.7070000171661377},{"id":"https://openalex.org/keywords/inference-efficiency","display_name":"inference efficiency","score":0.6349999904632568}],"concepts":[{"id":"https://openalex.org/C66322947","wikidata":"https://www.wikidata.org/wiki/Q11658","display_name":"Transformer","level":3,"score":0.7171000242233276},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6448000073432922},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.6290000081062317},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.4684999883174896},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.4138999879360199},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.37610000371932983},{"id":"https://openalex.org/C206729178","wikidata":"https://www.wikidata.org/wiki/Q2271896","display_name":"Scheduling (production processes)","level":2,"score":0.34950000047683716},{"id":"https://openalex.org/C67186912","wikidata":"https://www.wikidata.org/wiki/Q367664","display_name":"Data modeling","level":2,"score":0.3359000086784363},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.31769999861717224},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.2944999933242798},{"id":"https://openalex.org/C160920958","wikidata":"https://www.wikidata.org/wiki/Q7662746","display_name":"Synthetic data","level":2,"score":0.2849000096321106},{"id":"https://openalex.org/C2776359362","wikidata":"https://www.wikidata.org/wiki/Q2145286","display_name":"Representation (politics)","level":3,"score":0.28060001134872437},{"id":"https://openalex.org/C113775141","wikidata":"https://www.wikidata.org/wiki/Q428691","display_name":"Computer engineering","level":1,"score":0.28029999136924744},{"id":"https://openalex.org/C45374587","wikidata":"https://www.wikidata.org/wiki/Q12525525","display_name":"Computation","level":2,"score":0.2736000120639801},{"id":"https://openalex.org/C127413603","wikidata":"https://www.wikidata.org/wiki/Q11023","display_name":"Engineering","level":0,"score":0.2694999873638153},{"id":"https://openalex.org/C99844830","wikidata":"https://www.wikidata.org/wiki/Q102441924","display_name":"Scaling","level":2,"score":0.26669999957084656},{"id":"https://openalex.org/C79403827","wikidata":"https://www.wikidata.org/wiki/Q3988","display_name":"Real-time computing","level":1,"score":0.26600000262260437}],"mesh":[],"locations_count":1,"locations":[{"id":"pmh:oai:DiVA.org:kth-365780","doi":null,"is_oa":true,"landing_page_url":"http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-365780","pdf_url":"https://kth.diva-portal.org/smash/get/diva2:1979081/FULLTEXT01","source":{"id":"https://openalex.org/S4306401559","display_name":"KTH Publication Database DiVA (KTH Royal Institute of Technology)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I86987016","host_organization_name":"KTH Royal Institute of Technology","host_organization_lineage":["https://openalex.org/I86987016"],"host_organization_lineage_names":[],"type":"repository"},"license":"other-oa","license_id":"https://openalex.org/licenses/other-oa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"info:eu-repo/semantics/bachelorThesis"}],"best_oa_location":{"id":"pmh:oai:DiVA.org:kth-365780","doi":null,"is_oa":true,"landing_page_url":"http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-365780","pdf_url":"https://kth.diva-portal.org/smash/get/diva2:1979081/FULLTEXT01","source":{"id":"https://openalex.org/S4306401559","display_name":"KTH Publication Database DiVA (KTH Royal Institute of Technology)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I86987016","host_organization_name":"KTH Royal Institute of Technology","host_organization_lineage":["https://openalex.org/I86987016"],"host_organization_lineage_names":[],"type":"repository"},"license":"other-oa","license_id":"https://openalex.org/licenses/other-oa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"info:eu-repo/semantics/bachelorThesis"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7047635990.pdf","grobid_xml":"https://content.openalex.org/works/W7047635990.grobid-xml"},"referenced_works_count":28,"referenced_works":["https://openalex.org/W1498436455","https://openalex.org/W1614298861","https://openalex.org/W1810943226","https://openalex.org/W2130942839","https://openalex.org/W2257979135","https://openalex.org/W2626778328","https://openalex.org/W2736601468","https://openalex.org/W2767286248","https://openalex.org/W2896457183","https://openalex.org/W2899663614","https://openalex.org/W2908510526","https://openalex.org/W2964121744","https://openalex.org/W3001279689","https://openalex.org/W3030163527","https://openalex.org/W3033529678","https://openalex.org/W3094502228","https://openalex.org/W3119866685","https://openalex.org/W3188324247","https://openalex.org/W4224035735","https://openalex.org/W4226278401","https://openalex.org/W4386142098","https://openalex.org/W4390189088","https://openalex.org/W4401306886","https://openalex.org/W4405903187","https://openalex.org/W4413019822","https://openalex.org/W4415797729","https://openalex.org/W4415800382","https://openalex.org/W7133214597"],"related_works":["https://openalex.org/W4387596623","https://openalex.org/W3008851394","https://openalex.org/W7035948810","https://openalex.org/W7222241426","https://openalex.org/W4404569111","https://openalex.org/W3034340181","https://openalex.org/W4403345774","https://openalex.org/W4385822522","https://openalex.org/W4387010609","https://openalex.org/W4319995060"],"abstract_inverted_index":{"In":[0],"recent":[1],"years,":[2],"transformers":[3,76],"have":[4],"revolutionized":[5],"machine":[6],"learning,":[7],"catalyzing":[8],"huge":[9],"performance":[10,22],"gains":[11],"across":[12,104],"multiple":[13],"disciplines.":[14],"Due":[15],"to":[16,20,35,90,120,150],"the":[17,31,88,92,98,118],"vast":[18],"benefits":[19],"transformer":[21,105],"of":[23,53,58,72,87,100],"scaling":[24],"up":[25],"training":[26,46,73],"data":[27],"and":[28,37,47,55,63,74,125],"model":[29,89,109,112],"size,":[30],"computational":[32],"power":[33],"needed":[34],"train":[36],"deploy":[38],"modern":[39],"state-of-the-art":[40],"models":[41,50],"has":[42,117],"increased":[43],"drastically.":[44],"Today,":[45],"deploying":[48,75],"frontier":[49],"costs":[51],"millions":[52],"dollars":[54],"consumes":[56],"terawatt-hours":[57],"energy,":[59],"raising":[60],"both":[61],"financial":[62],"environmental":[64],"concerns.":[65],"This":[66],"report":[67],"explores":[68],"a":[69,137],"novel":[70],"method":[71,116],"by":[77],"using":[78],"Matryoshka":[79],"Representation":[80],"Learning":[81],"(MRL)":[82],"in":[83],"all":[84],"major":[85],"components":[86],"improve":[91],"compute-performance":[93],"tradeoff.":[94],"We":[95],"show":[96],"how":[97],"introduction":[99],"appropriate":[101],"coarse-to-fine":[102],"granularities":[103],"blocks":[106],"efficiently":[107],"reduces":[108],"parameters,":[110],"optimizing":[111],"size.":[113],"The":[114],"proposed":[115],"potential":[119],"significantly":[121],"reduce":[122],"inference":[123],"time":[124],"resource":[126],"use":[127],"while":[128],"causing":[129],"minimal":[130],"overhead":[131],"during":[132],"training.":[133],"Furthermore,":[134],"we":[135],"outline":[136],"dynamic":[138],"nesting":[139],"strategy":[140],"for":[141,144],"elastic":[142],"inference,":[143],"which":[145],"empirical":[146],"validation":[147],"is":[148],"left":[149],"future":[151],"research.":[152]},"counts_by_year":[],"updated_date":"2026-11-07T11:25:07.637180","created_date":"2025-10-10T00:00:00"}