{"id":"https://openalex.org/W7130654318","doi":"https://doi.org/10.1142/s1793351x26410023","title":"Agent Fine\u2013Tuning with Knowledge Distillation: Achieving Cost\u2013Efficient and High\u2013Fidelity LLMs in Microdomains","display_name":"Agent Fine\u2013Tuning with Knowledge Distillation: Achieving Cost\u2013Efficient and High\u2013Fidelity LLMs in Microdomains","publication_year":2026,"publication_date":"2026-02-20","ids":{"openalex":"https://openalex.org/W7130654318","doi":"https://doi.org/10.1142/s1793351x26410023"},"language":"en","primary_location":{"id":"doi:10.1142/s1793351x26410023","is_oa":false,"landing_page_url":"https://doi.org/10.1142/s1793351x26410023","pdf_url":null,"source":{"id":"https://openalex.org/S4210201727","display_name":"International Journal of Semantic Computing","issn_l":"1793-351X","issn":["1793-351X","1793-7108"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319815","host_organization_name":"World Scientific","host_organization_lineage":["https://openalex.org/P4310319815"],"host_organization_lineage_names":["World Scientific"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"International Journal of Semantic Computing","raw_type":"journal-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":false,"oa_status":"closed","oa_url":null,"any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5101012633","display_name":"Yawen Xue","orcid":null},"institutions":[{"id":"https://openalex.org/I65143321","display_name":"Hitachi (Japan)","ror":"https://ror.org/02exqgm79","country_code":"JP","type":"company","lineage":["https://openalex.org/I65143321"]}],"countries":["JP"],"is_corresponding":false,"raw_author_name":"Yawen Xue","raw_affiliation_strings":["Research and Development Group, Hitachi, Ltd., Tokyo, Japan"],"raw_orcid":"https://orcid.org/0009-0006-8173-0153","affiliations":[{"raw_affiliation_string":"Research and Development Group, Hitachi, Ltd., Tokyo, Japan","institution_ids":["https://openalex.org/I65143321"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5126460069","display_name":"Masaya Tsunokake","orcid":null},"institutions":[{"id":"https://openalex.org/I65143321","display_name":"Hitachi (Japan)","ror":"https://ror.org/02exqgm79","country_code":"JP","type":"company","lineage":["https://openalex.org/I65143321"]}],"countries":["JP"],"is_corresponding":false,"raw_author_name":"Masaya Tsunokake","raw_affiliation_strings":["Research and Development Group, Hitachi, Ltd., Tokyo, Japan"],"raw_orcid":"https://orcid.org/0000-0002-9885-7584","affiliations":[{"raw_affiliation_string":"Research and Development Group, Hitachi, Ltd., Tokyo, Japan","institution_ids":["https://openalex.org/I65143321"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5126453914","display_name":"Yuta Koreeda","orcid":null},"institutions":[{"id":"https://openalex.org/I65143321","display_name":"Hitachi (Japan)","ror":"https://ror.org/02exqgm79","country_code":"JP","type":"company","lineage":["https://openalex.org/I65143321"]}],"countries":["JP"],"is_corresponding":false,"raw_author_name":"Yuta Koreeda","raw_affiliation_strings":["Research and Development Group, Hitachi, Ltd., Tokyo, Japan"],"raw_orcid":"https://orcid.org/0009-0007-2262-3072","affiliations":[{"raw_affiliation_string":"Research and Development Group, Hitachi, Ltd., Tokyo, Japan","institution_ids":["https://openalex.org/I65143321"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5038587308","display_name":"Yasuhiro Sogawa","orcid":null},"institutions":[{"id":"https://openalex.org/I65143321","display_name":"Hitachi (Japan)","ror":"https://ror.org/02exqgm79","country_code":"JP","type":"company","lineage":["https://openalex.org/I65143321"]}],"countries":["JP"],"is_corresponding":false,"raw_author_name":"Yasuhiro Sogawa","raw_affiliation_strings":["Research and Development Group, Hitachi, Ltd., Tokyo, Japan"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Research and Development Group, Hitachi, Ltd., Tokyo, Japan","institution_ids":["https://openalex.org/I65143321"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I65143321"],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.20327789,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":"20","issue":"01","first_page":"31","last_page":"44"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.5019999742507935,"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":0.5019999742507935,"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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.09160000085830688,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.04740000143647194,"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/scalability","display_name":"Scalability","score":0.5181000232696533},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.5026999711990356},{"id":"https://openalex.org/keywords/distillation","display_name":"Distillation","score":0.4691999852657318},{"id":"https://openalex.org/keywords/knowledge-base","display_name":"Knowledge base","score":0.42730000615119934},{"id":"https://openalex.org/keywords/matching","display_name":"Matching (statistics)","score":0.41510000824928284},{"id":"https://openalex.org/keywords/work","display_name":"Work (physics)","score":0.36730000376701355}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7059000134468079},{"id":"https://openalex.org/C48044578","wikidata":"https://www.wikidata.org/wiki/Q727490","display_name":"Scalability","level":2,"score":0.5181000232696533},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.5026999711990356},{"id":"https://openalex.org/C204030448","wikidata":"https://www.wikidata.org/wiki/Q101017","display_name":"Distillation","level":2,"score":0.4691999852657318},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.43050000071525574},{"id":"https://openalex.org/C4554734","wikidata":"https://www.wikidata.org/wiki/Q593744","display_name":"Knowledge base","level":2,"score":0.42730000615119934},{"id":"https://openalex.org/C165064840","wikidata":"https://www.wikidata.org/wiki/Q1321061","display_name":"Matching (statistics)","level":2,"score":0.41510000824928284},{"id":"https://openalex.org/C18762648","wikidata":"https://www.wikidata.org/wiki/Q42213","display_name":"Work (physics)","level":2,"score":0.36730000376701355},{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.3142000138759613},{"id":"https://openalex.org/C117978034","wikidata":"https://www.wikidata.org/wiki/Q5422192","display_name":"Extractor","level":2,"score":0.30160000920295715},{"id":"https://openalex.org/C112930515","wikidata":"https://www.wikidata.org/wiki/Q4389547","display_name":"Risk analysis (engineering)","level":1,"score":0.2962999939918518},{"id":"https://openalex.org/C77967617","wikidata":"https://www.wikidata.org/wiki/Q4677561","display_name":"Active learning (machine learning)","level":2,"score":0.28459998965263367},{"id":"https://openalex.org/C9233905","wikidata":"https://www.wikidata.org/wiki/Q3276328","display_name":"Bidding","level":2,"score":0.2840999960899353},{"id":"https://openalex.org/C46304622","wikidata":"https://www.wikidata.org/wiki/Q374814","display_name":"Certification","level":2,"score":0.2705000042915344}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1142/s1793351x26410023","is_oa":false,"landing_page_url":"https://doi.org/10.1142/s1793351x26410023","pdf_url":null,"source":{"id":"https://openalex.org/S4210201727","display_name":"International Journal of Semantic Computing","issn_l":"1793-351X","issn":["1793-351X","1793-7108"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319815","host_organization_name":"World Scientific","host_organization_lineage":["https://openalex.org/P4310319815"],"host_organization_lineage_names":["World Scientific"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"International Journal of Semantic Computing","raw_type":"journal-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":8,"referenced_works":["https://openalex.org/W3177765786","https://openalex.org/W4391876619","https://openalex.org/W4392366650","https://openalex.org/W4402671772","https://openalex.org/W4402672016","https://openalex.org/W4402683785","https://openalex.org/W4405079446","https://openalex.org/W4409157490"],"related_works":[],"abstract_inverted_index":{"Agentic":[0],"large":[1],"language":[2],"models":[3,54,104],"(LLMs)":[4],"have":[5],"emerged":[6],"as":[7,198],"powerful":[8],"tools":[9],"for":[10,87,218],"autonomously":[11],"interacting":[12],"with":[13,28,84,114,134,165,207],"external":[14],"environments":[15],"and":[16,44,60,65,111,119,138,147,162,180,214],"performing":[17],"multi-step":[18],"reasoning.":[19],"While":[20],"most":[21],"existing":[22],"approaches":[23],"rely":[24],"on":[25,98,159],"in-context":[26],"learning":[27],"multi-turn":[29],"few-shot":[30],"prompts,":[31],"these":[32],"methods":[33],"often":[34],"require":[35],"long":[36],"inputs":[37],"and,":[38],"consequently,":[39],"incur":[40],"high":[41],"computational":[42],"costs":[43],"latency.":[45],"Agent":[46],"fine-tuning":[47,83,205],"offers":[48],"a":[49,94,140,211],"resource-aware":[50],"alternative":[51],"by":[52,122],"enabling":[53],"to":[55,90,144,187,224],"internalize":[56],"procedural":[57],"reasoning":[58,116],"patterns":[59],"domain-specific":[61],"knowledge":[62,85,208],"through":[63],"demonstrations":[64],"curated":[66],"training":[67],"data.":[68],"However,":[69],"its":[70],"effectiveness":[71],"in":[72],"highly":[73,212],"specialized":[74,225],"technical":[75,226],"microdomains":[76],"remains":[77],"underexplored.":[78],"This":[79],"work":[80],"investigates":[81],"agent":[82,204],"distillation":[86,209],"adapting":[88],"LLMs":[89,124,196],"Hitachi\u2019s":[91],"JP1":[92,150,160],"middleware,":[93],"complex":[95],"microdomain":[96],"centered":[97],"IT":[99],"operations":[100],"management.":[101],"We":[102],"fine-tune":[103],"using":[105],"JP1-specific":[106],"corpora":[107],"extracted":[108],"from":[109],"manuals":[110,161],"textbooks,":[112],"together":[113],"distilled":[115],"trajectories":[117,167],"(ReAct":[118],"CoT)":[120],"generated":[121],"larger":[123],"(GPT-4).":[125],"At":[126],"inference":[127],"time,":[128],"we":[129],"incorporate":[130],"retrieval-augmented":[131],"generation":[132],"(RAG)":[133],"an":[135],"agentic":[136],"prompt":[137],"introduce":[139],"context-answer":[141],"extractor":[142],"(CAE)":[143],"improve":[145],"grounding":[146],"relevance.":[148],"On":[149],"certification":[151],"examinations,":[152],"our":[153],"model":[154,174],"that":[155,203],"was":[156],"continually":[157],"pre-trained":[158],"further":[163],"fine-tuned":[164],"ReAct":[166],"achieves":[168],"substantial":[169],"improvements":[170],"over":[171],"the":[172],"base":[173],"\u2014":[175,183],"13%":[176],"(Engineer),":[177],"12%":[178],"(Professional),":[179],"10%":[181],"(Consultant)":[182],"while":[184],"delivering":[185],"up":[186],"9.78":[188],"times":[189],"higher":[190],"cost":[191],"efficiency":[192],"than":[193],"strong":[194],"general-purpose":[195],"such":[197],"GPT-4.":[199],"These":[200],"results":[201],"demonstrate":[202],"combined":[206],"is":[210],"effective":[213],"economically":[215],"scalable":[216],"strategy":[217],"building":[219],"high-fidelity":[220],"LLM":[221],"agents":[222],"tailored":[223],"domains.":[227]},"counts_by_year":[],"updated_date":"2026-06-26T08:34:08.712188","created_date":"2026-02-20T00:00:00"}