{"id":"https://openalex.org/W7129409960","doi":"https://doi.org/10.3389/frai.2026.1665992","title":"An efficient strategy for fine-tuning large language models","display_name":"An efficient strategy for fine-tuning large language models","publication_year":2026,"publication_date":"2026-02-17","ids":{"openalex":"https://openalex.org/W7129409960","doi":"https://doi.org/10.3389/frai.2026.1665992","pmid":"https://pubmed.ncbi.nlm.nih.gov/41783256"},"language":"en","primary_location":{"id":"doi:10.3389/frai.2026.1665992","is_oa":true,"landing_page_url":"https://doi.org/10.3389/frai.2026.1665992","pdf_url":"https://public-pages-files-2025.frontiersin.org/journals/artificial-intelligence/articles/10.3389/frai.2026.1665992/pdf","source":{"id":"https://openalex.org/S4210197006","display_name":"Frontiers in Artificial Intelligence","issn_l":"2624-8212","issn":["2624-8212"],"is_oa":true,"is_in_doaj":true,"is_core":true,"host_organization":"https://openalex.org/P4310320527","host_organization_name":"Frontiers Media","host_organization_lineage":["https://openalex.org/P4310320527"],"host_organization_lineage_names":["Frontiers Media"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Frontiers in Artificial Intelligence","raw_type":"journal-article"},"type":"article","indexed_in":["crossref","doaj","pubmed"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://public-pages-files-2025.frontiersin.org/journals/artificial-intelligence/articles/10.3389/frai.2026.1665992/pdf","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5038454024","display_name":"Benjamin M. Marsh","orcid":null},"institutions":[{"id":"https://openalex.org/I4210162473","display_name":"United States Marine Corps","ror":"https://ror.org/05av6m066","country_code":"US","type":"government","lineage":["https://openalex.org/I1330347796","https://openalex.org/I3130687028","https://openalex.org/I4210162473"]}],"countries":["US"],"is_corresponding":true,"raw_author_name":"Benjamin Marsh","raw_affiliation_strings":["Marine Corps Tactical Systems Support Activity, United States Marine Corps"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Marine Corps Tactical Systems Support Activity, United States Marine Corps","institution_ids":["https://openalex.org/I4210162473"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5016624503","display_name":"Adam Michaleas","orcid":"https://orcid.org/0000-0001-7402-8303"},"institutions":[{"id":"https://openalex.org/I4210122954","display_name":"MIT Lincoln Laboratory","ror":"https://ror.org/022z6jk58","country_code":"US","type":"facility","lineage":["https://openalex.org/I4210122954","https://openalex.org/I63966007"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Adam Michaleas","raw_affiliation_strings":["MIT Lincoln Laboratory, Artificial Intelligence Technology"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"MIT Lincoln Laboratory, Artificial Intelligence Technology","institution_ids":["https://openalex.org/I4210122954"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5031850946","display_name":"Darrell Ricke","orcid":"https://orcid.org/0000-0002-2842-2809"},"institutions":[{"id":"https://openalex.org/I4210122954","display_name":"MIT Lincoln Laboratory","ror":"https://ror.org/022z6jk58","country_code":"US","type":"facility","lineage":["https://openalex.org/I4210122954","https://openalex.org/I63966007"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Darrell O. Ricke","raw_affiliation_strings":["MIT Lincoln Laboratory, Artificial Intelligence Technology"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"MIT Lincoln Laboratory, Artificial Intelligence Technology","institution_ids":["https://openalex.org/I4210122954"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5054337078","display_name":"Shaun Monera","orcid":null},"institutions":[{"id":"https://openalex.org/I4210162473","display_name":"United States Marine Corps","ror":"https://ror.org/05av6m066","country_code":"US","type":"government","lineage":["https://openalex.org/I1330347796","https://openalex.org/I3130687028","https://openalex.org/I4210162473"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Shaun Monera","raw_affiliation_strings":["Marine Corps Tactical Systems Support Activity, United States Marine Corps"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Marine Corps Tactical Systems Support Activity, United States Marine Corps","institution_ids":["https://openalex.org/I4210162473"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5126214804","display_name":"Shriya Zembruski","orcid":null},"institutions":[{"id":"https://openalex.org/I4210162473","display_name":"United States Marine Corps","ror":"https://ror.org/05av6m066","country_code":"US","type":"government","lineage":["https://openalex.org/I1330347796","https://openalex.org/I3130687028","https://openalex.org/I4210162473"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Shriya Zembruski","raw_affiliation_strings":["Marine Corps Tactical Systems Support Activity, United States Marine Corps"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Marine Corps Tactical Systems Support Activity, United States Marine Corps","institution_ids":["https://openalex.org/I4210162473"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":2,"corresponding_author_ids":["https://openalex.org/A5038454024"],"corresponding_institution_ids":["https://openalex.org/I4210162473"],"apc_list":{"value":2125,"currency":"USD","value_usd":2125},"apc_paid":{"value":2125,"currency":"USD","value_usd":2125},"fwci":7.296,"has_fulltext":true,"cited_by_count":1,"citation_normalized_percentile":{"value":0.94887284,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":94,"max":97},"biblio":{"volume":"9","issue":null,"first_page":"1665992","last_page":"1665992"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.37220001220703125,"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/T10028","display_name":"Topic Modeling","score":0.37220001220703125,"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/T12535","display_name":"Machine Learning and Data Classification","score":0.07999999821186066,"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/T11636","display_name":"Artificial Intelligence in Healthcare and Education","score":0.05689999833703041,"subfield":{"id":"https://openalex.org/subfields/2718","display_name":"Health Informatics"},"field":{"id":"https://openalex.org/fields/27","display_name":"Medicine"},"domain":{"id":"https://openalex.org/domains/4","display_name":"Health Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/workflow","display_name":"Workflow","score":0.6158000230789185},{"id":"https://openalex.org/keywords/benchmark","display_name":"Benchmark (surveying)","score":0.6136000156402588},{"id":"https://openalex.org/keywords/process","display_name":"Process (computing)","score":0.5127000212669373},{"id":"https://openalex.org/keywords/task","display_name":"Task (project management)","score":0.4869999885559082},{"id":"https://openalex.org/keywords/adaptation","display_name":"Adaptation (eye)","score":0.48249998688697815},{"id":"https://openalex.org/keywords/domain","display_name":"Domain (mathematical analysis)","score":0.4251999855041504},{"id":"https://openalex.org/keywords/modalities","display_name":"Modalities","score":0.41269999742507935},{"id":"https://openalex.org/keywords/language-model","display_name":"Language model","score":0.40310001373291016},{"id":"https://openalex.org/keywords/construct","display_name":"Construct (python library)","score":0.39010000228881836}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8184999823570251},{"id":"https://openalex.org/C177212765","wikidata":"https://www.wikidata.org/wiki/Q627335","display_name":"Workflow","level":2,"score":0.6158000230789185},{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.6136000156402588},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.5382000207901001},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.5127000212669373},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.4869999885559082},{"id":"https://openalex.org/C139807058","wikidata":"https://www.wikidata.org/wiki/Q352374","display_name":"Adaptation (eye)","level":2,"score":0.48249998688697815},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.462799996137619},{"id":"https://openalex.org/C36503486","wikidata":"https://www.wikidata.org/wiki/Q11235244","display_name":"Domain (mathematical analysis)","level":2,"score":0.4251999855041504},{"id":"https://openalex.org/C2779903281","wikidata":"https://www.wikidata.org/wiki/Q6888026","display_name":"Modalities","level":2,"score":0.41269999742507935},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.40310001373291016},{"id":"https://openalex.org/C2780801425","wikidata":"https://www.wikidata.org/wiki/Q5164392","display_name":"Construct (python library)","level":2,"score":0.39010000228881836},{"id":"https://openalex.org/C2780226545","wikidata":"https://www.wikidata.org/wiki/Q6888030","display_name":"Modality (human\u2013computer interaction)","level":2,"score":0.34940001368522644},{"id":"https://openalex.org/C206345919","wikidata":"https://www.wikidata.org/wiki/Q20380951","display_name":"Resource (disambiguation)","level":2,"score":0.3492000102996826},{"id":"https://openalex.org/C195324797","wikidata":"https://www.wikidata.org/wiki/Q33742","display_name":"Natural language","level":2,"score":0.34689998626708984},{"id":"https://openalex.org/C206729178","wikidata":"https://www.wikidata.org/wiki/Q2271896","display_name":"Scheduling (production processes)","level":2,"score":0.33880001306533813},{"id":"https://openalex.org/C2779439875","wikidata":"https://www.wikidata.org/wiki/Q1078276","display_name":"Natural language understanding","level":3,"score":0.32739999890327454},{"id":"https://openalex.org/C165064840","wikidata":"https://www.wikidata.org/wiki/Q1321061","display_name":"Matching (statistics)","level":2,"score":0.3215999901294708},{"id":"https://openalex.org/C2164484","wikidata":"https://www.wikidata.org/wiki/Q5170150","display_name":"Core (optical fiber)","level":2,"score":0.3156999945640564},{"id":"https://openalex.org/C29202148","wikidata":"https://www.wikidata.org/wiki/Q287260","display_name":"Resource allocation","level":2,"score":0.3154999911785126},{"id":"https://openalex.org/C175154964","wikidata":"https://www.wikidata.org/wiki/Q380077","display_name":"Task analysis","level":3,"score":0.3066999912261963},{"id":"https://openalex.org/C86251818","wikidata":"https://www.wikidata.org/wiki/Q816754","display_name":"Benchmarking","level":2,"score":0.2987000048160553},{"id":"https://openalex.org/C67186912","wikidata":"https://www.wikidata.org/wiki/Q367664","display_name":"Data modeling","level":2,"score":0.290800005197525},{"id":"https://openalex.org/C2777212361","wikidata":"https://www.wikidata.org/wiki/Q5127848","display_name":"Class (philosophy)","level":2,"score":0.287200003862381},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.2632000148296356},{"id":"https://openalex.org/C774472","wikidata":"https://www.wikidata.org/wiki/Q6760393","display_name":"Margin (machine learning)","level":2,"score":0.260699987411499},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.2590000033378601},{"id":"https://openalex.org/C2776187449","wikidata":"https://www.wikidata.org/wiki/Q1513879","display_name":"Natural language generation","level":3,"score":0.25690001249313354},{"id":"https://openalex.org/C2781215313","wikidata":"https://www.wikidata.org/wiki/Q3493345","display_name":"SPARK (programming language)","level":2,"score":0.2565999925136566}],"mesh":[],"locations_count":4,"locations":[{"id":"doi:10.3389/frai.2026.1665992","is_oa":true,"landing_page_url":"https://doi.org/10.3389/frai.2026.1665992","pdf_url":"https://public-pages-files-2025.frontiersin.org/journals/artificial-intelligence/articles/10.3389/frai.2026.1665992/pdf","source":{"id":"https://openalex.org/S4210197006","display_name":"Frontiers in Artificial Intelligence","issn_l":"2624-8212","issn":["2624-8212"],"is_oa":true,"is_in_doaj":true,"is_core":true,"host_organization":"https://openalex.org/P4310320527","host_organization_name":"Frontiers Media","host_organization_lineage":["https://openalex.org/P4310320527"],"host_organization_lineage_names":["Frontiers Media"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Frontiers in Artificial Intelligence","raw_type":"journal-article"},{"id":"pmid:41783256","is_oa":false,"landing_page_url":"https://pubmed.ncbi.nlm.nih.gov/41783256","pdf_url":null,"source":{"id":"https://openalex.org/S4306525036","display_name":"PubMed","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I1299303238","host_organization_name":"National Institutes of Health","host_organization_lineage":["https://openalex.org/I1299303238"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Frontiers in artificial intelligence","raw_type":"Journal Article"},{"id":"pmh:oai:doaj.org/article:b17fc2ee0b3c4dabbd65ea3f7aa1042c","is_oa":true,"landing_page_url":"https://doaj.org/article/b17fc2ee0b3c4dabbd65ea3f7aa1042c","pdf_url":null,"source":{"id":"https://openalex.org/S4306401280","display_name":"DOAJ (DOAJ: Directory of Open Access Journals)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by-sa","license_id":"https://openalex.org/licenses/cc-by-sa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"Frontiers in Artificial Intelligence, Vol 9 (2026)","raw_type":"article"},{"id":"pmh:oai:pubmedcentral.nih.gov:12953369","is_oa":true,"landing_page_url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC12953369/","pdf_url":null,"source":{"id":"https://openalex.org/S2764455111","display_name":"PubMed Central","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I1299303238","host_organization_name":"National Institutes of Health","host_organization_lineage":["https://openalex.org/I1299303238"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"Front Artif Intell","raw_type":"Text"}],"best_oa_location":{"id":"doi:10.3389/frai.2026.1665992","is_oa":true,"landing_page_url":"https://doi.org/10.3389/frai.2026.1665992","pdf_url":"https://public-pages-files-2025.frontiersin.org/journals/artificial-intelligence/articles/10.3389/frai.2026.1665992/pdf","source":{"id":"https://openalex.org/S4210197006","display_name":"Frontiers in Artificial Intelligence","issn_l":"2624-8212","issn":["2624-8212"],"is_oa":true,"is_in_doaj":true,"is_core":true,"host_organization":"https://openalex.org/P4310320527","host_organization_name":"Frontiers Media","host_organization_lineage":["https://openalex.org/P4310320527"],"host_organization_lineage_names":["Frontiers Media"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Frontiers in Artificial 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/W7129409960.pdf","grobid_xml":"https://content.openalex.org/works/W7129409960.grobid-xml"},"referenced_works_count":16,"referenced_works":["https://openalex.org/W2890431379","https://openalex.org/W3034238904","https://openalex.org/W3129831491","https://openalex.org/W3214600982","https://openalex.org/W4221143046","https://openalex.org/W4288089799","https://openalex.org/W4360891289","https://openalex.org/W4365511667","https://openalex.org/W4376122773","https://openalex.org/W4385245566","https://openalex.org/W4385571011","https://openalex.org/W4386501849","https://openalex.org/W4393138529","https://openalex.org/W4406054803","https://openalex.org/W4408793641","https://openalex.org/W4413237296"],"related_works":[],"abstract_inverted_index":{"Introduction:":[0],"Large":[1],"Language":[2,11],"Models":[3],"(LLMs)":[4],"achieve":[5],"strong":[6],"performance":[7,188],"on":[8,219],"many":[9],"Natural":[10],"Processing":[12],"tasks,":[13],"but":[14],"adapting":[15],"them":[16],"to":[17,23,209],"domain-specific":[18,47,242],"applications":[19],"is":[20],"resource-intensive":[21],"due":[22],"the":[24,31,89,111,135,144,175,193,215],"cost":[25],"of":[26,113,181,187],"curating":[27],"task-specific":[28],"datasets":[29],"and":[30,52,66,76,105,159,189],"compute":[32,53,190,221],"required":[33],"for":[34,42,46,63,82,203,237,241],"fine-tuning.":[35],"This":[36],"work":[37],"proposes":[38],"an":[39,119,155,178],"end-to-end":[40],"strategy":[41,58],"rapidly":[43],"fine-tuning":[44,95,142,216,239],"LLMs":[45,240],"tasks":[48,243],"when":[49,223,228],"both":[50],"data":[51,246],"are":[54],"limited.":[55],"Methods:":[56],"The":[57,230],"uses":[59],"Distilling":[60],"Step-by-Step":[61],"(DSS)":[62],"dataset":[64],"development":[65],"model":[67,72],"training,":[68],"where":[69],"a":[70,83,130,184,200,234],"teacher":[71],"generates":[73],"task":[74],"labels":[75],"intermediate":[77],"rationales":[78],"via":[79],"Chain-of-Thought":[80],"prompting":[81],"natural-language-to-Query-DSL":[84],"structured":[85],"generation":[86],"task.":[87],"Using":[88],"resulting":[90],"supervision,":[91,115],"we":[92,116],"benchmark":[93],"three":[94],"modalities":[96],"through":[97],"hyperparameter":[98],"sweeps:":[99],"full-precision":[100,141],"fine-tuning,":[101],"Low-Rank":[102],"Adaptation":[103],"(LoRA),":[104],"Quantized":[106],"LoRA":[107,153,225],"(QLoRA).":[108],"To":[109],"isolate":[110],"effect":[112],"rationale":[114,127],"additionally":[117],"conduct":[118],"ablation":[120],"study":[121],"comparing":[122],"DSS":[123,138,151,160,208],"training":[124,164],"(label":[125],"+":[126],"supervision)":[128],"against":[129],"label-only":[131],"configuration.":[132],"Results:":[133],"Across":[134],"evaluated":[136],"configurations,":[137],"combined":[139],"with":[140,152,161,244],"yields":[143],"strongest":[145],"overall":[146],"performance.":[147,173],"Under":[148],"resource":[149],"constraints,":[150],"provides":[154,183],"effective":[156],"performance-efficiency":[157],"tradeoff,":[158],"QLoRA":[162,227],"enables":[163],"under":[165],"tighter":[166],"GPU":[167],"memory":[168],"budgets":[169],"while":[170],"maintaining":[171],"competitive":[172],"In":[174],"parameter-efficient":[176],"regimes,":[177],"alpha-to-rank":[179],"ratio":[180],"4:1":[182],"consistent":[185],"balance":[186],"consumption":[191],"across":[192],"explored":[194],"settings.":[195],"Discussion:":[196],"These":[197],"findings":[198],"support":[199],"practical":[201],"process":[202],"resource-constrained":[204],"domain":[205],"adaptation:":[206],"use":[207],"efficiently":[210,238],"construct":[211],"datasets,":[212],"then":[213],"select":[214],"modality":[217],"based":[218],"available":[220],"(full-precision":[222],"feasible;":[224],"or":[226],"memory-limited).":[229],"proposed":[231],"workflow":[232],"offers":[233],"general":[235],"guide":[236],"limited":[245],"availability.":[247]},"counts_by_year":[{"year":2026,"cited_by_count":1}],"updated_date":"2026-08-18T07:49:30.821534","created_date":"2026-02-18T00:00:00"}