{"id":"https://openalex.org/W4406844855","doi":"https://doi.org/10.1038/s42256-024-00975-8","title":"Evolutionary optimization of model merging recipes","display_name":"Evolutionary optimization of model merging recipes","publication_year":2025,"publication_date":"2025-01-27","ids":{"openalex":"https://openalex.org/W4406844855","doi":"https://doi.org/10.1038/s42256-024-00975-8"},"language":"en","primary_location":{"id":"doi:10.1038/s42256-024-00975-8","is_oa":true,"landing_page_url":"https://doi.org/10.1038/s42256-024-00975-8","pdf_url":"https://www.nature.com/articles/s42256-024-00975-8.pdf","source":{"id":"https://openalex.org/S2912241403","display_name":"Nature Machine Intelligence","issn_l":"2522-5839","issn":["2522-5839"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319908","host_organization_name":"Nature Portfolio","host_organization_lineage":["https://openalex.org/P4310319908","https://openalex.org/P4310319965"],"host_organization_lineage_names":["Nature Portfolio","Springer Nature"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Nature Machine Intelligence","raw_type":"journal-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"hybrid","oa_url":"https://www.nature.com/articles/s42256-024-00975-8.pdf","any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5094227651","display_name":"Takuya Akiba","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Takuya Akiba","raw_affiliation_strings":["Sakana AI, Tokyo, Japan"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Sakana AI, Tokyo, Japan","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5094227652","display_name":"Makoto Shing","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Makoto Shing","raw_affiliation_strings":["Sakana AI, Tokyo, Japan"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Sakana AI, Tokyo, Japan","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5094227653","display_name":"Yujin Tang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yujin Tang","raw_affiliation_strings":["Sakana AI, Tokyo, Japan"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Sakana AI, Tokyo, Japan","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5112226018","display_name":"Qi Sun","orcid":"https://orcid.org/0009-0004-9875-0906"},"institutions":[{"id":"https://openalex.org/I4400009020","display_name":"Institute of Science Tokyo","ror":"https://ror.org/05dqf9946","country_code":null,"type":"education","lineage":["https://openalex.org/I4400009020"]}],"countries":["JP"],"is_corresponding":false,"raw_author_name":"Qi Sun","raw_affiliation_strings":["Institute of Science Tokyo, Tokyo, Japan","Sakana AI, Tokyo, Japan"],"raw_orcid":"https://orcid.org/0009-0004-9875-0906","affiliations":[{"raw_affiliation_string":"Institute of Science Tokyo, Tokyo, Japan","institution_ids":["https://openalex.org/I4400009020"]},{"raw_affiliation_string":"Sakana AI, Tokyo, Japan","institution_ids":[]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5001419875","display_name":"David Ha","orcid":"https://orcid.org/0000-0001-8765-8574"},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"David Ha","raw_affiliation_strings":["Sakana AI, Tokyo, Japan"],"raw_orcid":"https://orcid.org/0000-0001-8765-8574","affiliations":[{"raw_affiliation_string":"Sakana AI, Tokyo, Japan","institution_ids":[]}]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":5,"corresponding_author_ids":["https://openalex.org/A5001419875"],"corresponding_institution_ids":[],"apc_list":{"value":9750,"currency":"EUR","value_usd":11690},"apc_paid":{"value":9750,"currency":"EUR","value_usd":11690},"fwci":79.0492,"has_fulltext":true,"cited_by_count":36,"citation_normalized_percentile":{"value":0.99939889,"is_in_top_1_percent":true,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":94,"max":100},"biblio":{"volume":"7","issue":"2","first_page":"195","last_page":"204"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10260","display_name":"Software Engineering Research","score":0.996399998664856,"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/T10260","display_name":"Software Engineering Research","score":0.996399998664856,"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/T10848","display_name":"Advanced Multi-Objective Optimization Algorithms","score":0.9908999800682068,"subfield":{"id":"https://openalex.org/subfields/1703","display_name":"Computational Theory and Mathematics"},"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/T11450","display_name":"Model-Driven Software Engineering Techniques","score":0.989300012588501,"subfield":{"id":"https://openalex.org/subfields/1712","display_name":"Software"},"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/computer-science","display_name":"Computer science","score":0.5099722146987915},{"id":"https://openalex.org/keywords/operations-research","display_name":"Operations research","score":0.3241501748561859},{"id":"https://openalex.org/keywords/business","display_name":"Business","score":0.3207321763038635},{"id":"https://openalex.org/keywords/engineering","display_name":"Engineering","score":0.14104056358337402}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5099722146987915},{"id":"https://openalex.org/C42475967","wikidata":"https://www.wikidata.org/wiki/Q194292","display_name":"Operations research","level":1,"score":0.3241501748561859},{"id":"https://openalex.org/C144133560","wikidata":"https://www.wikidata.org/wiki/Q4830453","display_name":"Business","level":0,"score":0.3207321763038635},{"id":"https://openalex.org/C127413603","wikidata":"https://www.wikidata.org/wiki/Q11023","display_name":"Engineering","level":0,"score":0.14104056358337402}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1038/s42256-024-00975-8","is_oa":true,"landing_page_url":"https://doi.org/10.1038/s42256-024-00975-8","pdf_url":"https://www.nature.com/articles/s42256-024-00975-8.pdf","source":{"id":"https://openalex.org/S2912241403","display_name":"Nature Machine Intelligence","issn_l":"2522-5839","issn":["2522-5839"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319908","host_organization_name":"Nature Portfolio","host_organization_lineage":["https://openalex.org/P4310319908","https://openalex.org/P4310319965"],"host_organization_lineage_names":["Nature Portfolio","Springer Nature"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Nature Machine Intelligence","raw_type":"journal-article"}],"best_oa_location":{"id":"doi:10.1038/s42256-024-00975-8","is_oa":true,"landing_page_url":"https://doi.org/10.1038/s42256-024-00975-8","pdf_url":"https://www.nature.com/articles/s42256-024-00975-8.pdf","source":{"id":"https://openalex.org/S2912241403","display_name":"Nature Machine Intelligence","issn_l":"2522-5839","issn":["2522-5839"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319908","host_organization_name":"Nature Portfolio","host_organization_lineage":["https://openalex.org/P4310319908","https://openalex.org/P4310319965"],"host_organization_lineage_names":["Nature Portfolio","Springer Nature"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Nature Machine Intelligence","raw_type":"journal-article"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4406844855.pdf","grobid_xml":"https://content.openalex.org/works/W4406844855.grobid-xml"},"referenced_works_count":18,"referenced_works":["https://openalex.org/W2089217417","https://openalex.org/W2126105956","https://openalex.org/W2563351168","https://openalex.org/W2949676527","https://openalex.org/W2963626623","https://openalex.org/W4221155125","https://openalex.org/W4225892134","https://openalex.org/W4303648904","https://openalex.org/W4376312115","https://openalex.org/W4385572928","https://openalex.org/W4386044080","https://openalex.org/W4386185600","https://openalex.org/W4402727764","https://openalex.org/W6803096969","https://openalex.org/W6847030267","https://openalex.org/W6858023062","https://openalex.org/W6949396384","https://openalex.org/W6968129167"],"related_works":["https://openalex.org/W4391375266","https://openalex.org/W2899084033","https://openalex.org/W2748952813","https://openalex.org/W2390279801","https://openalex.org/W4391913857","https://openalex.org/W2358668433","https://openalex.org/W4396701345","https://openalex.org/W2376932109","https://openalex.org/W2001405890","https://openalex.org/W4396696052"],"abstract_inverted_index":{"Abstract":[0],"Large":[1],"language":[2],"models":[3,30,106,134,178],"(LLMs)":[4],"have":[5],"become":[6],"increasingly":[7],"capable,":[8],"but":[9,184],"their":[10,67],"development":[11],"often":[12],"requires":[13],"substantial":[14],"computational":[15],"resources.":[16],"Although":[17],"model":[18,153,192,204],"merging":[19],"has":[20],"emerged":[21],"as":[22,108],"a":[23,109,125,148,187],"cost-effective":[24],"promising":[25],"approach":[26,52,79,100,157],"for":[27,144,190,197],"creating":[28],"new":[29,176,188],"by":[31,57],"combining":[32],"existing":[33],"ones,":[34],"it":[35],"currently":[36],"relies":[37],"on":[38,124],"human":[39],"intuition":[40],"and":[41,85],"domain":[42],"knowledge,":[43],"limiting":[44],"its":[45,159],"potential.":[46],"Here":[47],"we":[48],"propose":[49],"an":[50],"evolutionary":[51],"that":[53],"overcomes":[54],"this":[55],"limitation":[56],"automatically":[58],"discovering":[59],"effective":[60],"combinations":[61],"of":[62,95,127],"diverse":[63],"open-source":[64,182],"models,":[65],"harnessing":[66],"collective":[68],"intelligence":[69],"without":[70],"requiring":[71],"extensive":[72],"additional":[73],"training":[74],"data":[75,86],"or":[76],"compute.":[77],"Our":[78],"operates":[80],"in":[81,161],"both":[82],"parameter":[83],"space":[84],"flow":[87],"space,":[88],"allowing":[89],"optimization":[90],"beyond":[91],"just":[92],"the":[93,96,181,195],"weights":[94],"individual":[97],"models.":[98,170],"This":[99,171],"even":[101,132],"facilitates":[102],"cross-domain":[103],"merging,":[104],"generating":[105],"such":[107,145],"Japanese":[110,118,129,151,163,168],"LLM":[111,120,130],"with":[112,135],"math":[113,119],"reasoning":[114],"capabilities.":[115],"Surprisingly,":[116],"our":[117,156],"achieved":[121],"state-of-the-art":[122,177],"performance":[123],"variety":[126],"established":[128],"benchmarks,":[131],"surpassing":[133],"substantially":[136],"more":[137],"parameters,":[138],"despite":[139],"not":[140,173],"being":[141],"explicitly":[142],"trained":[143],"tasks.":[146],"Furthermore,":[147],"culturally":[149],"aware":[150],"vision\u2013language":[152,169],"generated":[154],"through":[155],"demonstrates":[158],"effectiveness":[160],"describing":[162],"culture-specific":[164],"content,":[165],"outperforming":[166],"previous":[167],"work":[172],"only":[174],"contributes":[175],"back":[179],"to":[180,202],"community":[183],"also":[185],"introduces":[186],"paradigm":[189],"automated":[191],"composition,":[193],"paving":[194],"way":[196],"exploring":[198],"alternative,":[199],"efficient":[200],"approaches":[201],"foundation":[203],"development.":[205]},"counts_by_year":[{"year":2026,"cited_by_count":9},{"year":2025,"cited_by_count":25},{"year":2024,"cited_by_count":2}],"updated_date":"2026-06-13T06:13:01.061226","created_date":"2025-10-10T00:00:00"}