{"id":"https://openalex.org/W4417534296","doi":"https://doi.org/10.48550/arxiv.2512.15943","title":"Small Language Models for Efficient Agentic Tool Calling: Outperforming Large Models with Targeted Fine-tuning","display_name":"Small Language Models for Efficient Agentic Tool Calling: Outperforming Large Models with Targeted Fine-tuning","publication_year":2025,"publication_date":"2025-12-17","ids":{"openalex":"https://openalex.org/W4417534296","doi":"https://doi.org/10.48550/arxiv.2512.15943"},"language":null,"primary_location":{"id":"pmh:oai:arXiv.org:2512.15943","is_oa":true,"landing_page_url":"http://arxiv.org/abs/2512.15943","pdf_url":"https://arxiv.org/pdf/2512.15943","source":{"id":"https://openalex.org/S4393918464","display_name":"ArXiv.org","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"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","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"type":"preprint","indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2512.15943","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":null,"display_name":"Jhandi, Polaris","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jhandi, Polaris","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Kazi, Owais","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Kazi, Owais","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Subramanian, Shreyas","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Subramanian, Shreyas","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":null,"display_name":"Sendas, Neel","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Sendas, Neel","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":4,"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":false,"primary_topic":{"id":"https://openalex.org/T11636","display_name":"Artificial Intelligence in Healthcare and Education","score":0.2547999918460846,"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"}},"topics":[{"id":"https://openalex.org/T11636","display_name":"Artificial Intelligence in Healthcare and Education","score":0.2547999918460846,"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"}},{"id":"https://openalex.org/T11948","display_name":"Machine Learning in Materials Science","score":0.1842000037431717,"subfield":{"id":"https://openalex.org/subfields/2505","display_name":"Materials Chemistry"},"field":{"id":"https://openalex.org/fields/25","display_name":"Materials Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.07450000196695328,"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/workflow","display_name":"Workflow","score":0.6432999968528748},{"id":"https://openalex.org/keywords/language-model","display_name":"Language model","score":0.6241999864578247},{"id":"https://openalex.org/keywords/scale","display_name":"Scale (ratio)","score":0.5038999915122986},{"id":"https://openalex.org/keywords/overhead","display_name":"Overhead (engineering)","score":0.4505999982357025},{"id":"https://openalex.org/keywords/generative-grammar","display_name":"Generative grammar","score":0.4251999855041504},{"id":"https://openalex.org/keywords/generative-model","display_name":"Generative model","score":0.40610000491142273},{"id":"https://openalex.org/keywords/reinforcement-learning","display_name":"Reinforcement learning","score":0.3788999915122986},{"id":"https://openalex.org/keywords/baseline","display_name":"Baseline (sea)","score":0.36579999327659607}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7724000215530396},{"id":"https://openalex.org/C177212765","wikidata":"https://www.wikidata.org/wiki/Q627335","display_name":"Workflow","level":2,"score":0.6432999968528748},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.6241999864578247},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5260999798774719},{"id":"https://openalex.org/C2778755073","wikidata":"https://www.wikidata.org/wiki/Q10858537","display_name":"Scale (ratio)","level":2,"score":0.5038999915122986},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.49160000681877136},{"id":"https://openalex.org/C2779960059","wikidata":"https://www.wikidata.org/wiki/Q7113681","display_name":"Overhead (engineering)","level":2,"score":0.4505999982357025},{"id":"https://openalex.org/C39890363","wikidata":"https://www.wikidata.org/wiki/Q36108","display_name":"Generative grammar","level":2,"score":0.4251999855041504},{"id":"https://openalex.org/C167966045","wikidata":"https://www.wikidata.org/wiki/Q5532625","display_name":"Generative model","level":3,"score":0.40610000491142273},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.3788999915122986},{"id":"https://openalex.org/C12725497","wikidata":"https://www.wikidata.org/wiki/Q810247","display_name":"Baseline (sea)","level":2,"score":0.36579999327659607},{"id":"https://openalex.org/C2779304628","wikidata":"https://www.wikidata.org/wiki/Q3503480","display_name":"Face (sociological concept)","level":2,"score":0.35030001401901245},{"id":"https://openalex.org/C93518851","wikidata":"https://www.wikidata.org/wiki/Q180160","display_name":"Metadata","level":2,"score":0.3197000026702881},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.31029999256134033},{"id":"https://openalex.org/C2777115002","wikidata":"https://www.wikidata.org/wiki/Q7168246","display_name":"Performance prediction","level":2,"score":0.30820000171661377},{"id":"https://openalex.org/C2778348673","wikidata":"https://www.wikidata.org/wiki/Q739302","display_name":"Production (economics)","level":2,"score":0.29789999127388},{"id":"https://openalex.org/C67186912","wikidata":"https://www.wikidata.org/wiki/Q367664","display_name":"Data modeling","level":2,"score":0.2924000024795532},{"id":"https://openalex.org/C184337299","wikidata":"https://www.wikidata.org/wiki/Q1437428","display_name":"Semantics (computer science)","level":2,"score":0.2854999899864197},{"id":"https://openalex.org/C45804977","wikidata":"https://www.wikidata.org/wiki/Q7239673","display_name":"Predictive modelling","level":2,"score":0.2784000039100647},{"id":"https://openalex.org/C108010975","wikidata":"https://www.wikidata.org/wiki/Q500094","display_name":"Pruning","level":2,"score":0.27559998631477356},{"id":"https://openalex.org/C206588197","wikidata":"https://www.wikidata.org/wiki/Q846574","display_name":"Reuse","level":2,"score":0.25049999356269836}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2512.15943","is_oa":true,"landing_page_url":"http://arxiv.org/abs/2512.15943","pdf_url":"https://arxiv.org/pdf/2512.15943","source":{"id":"https://openalex.org/S4393918464","display_name":"ArXiv.org","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"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","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2512.15943","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2512.15943","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":"pmh:oai:arXiv.org:2512.15943","is_oa":true,"landing_page_url":"http://arxiv.org/abs/2512.15943","pdf_url":"https://arxiv.org/pdf/2512.15943","source":{"id":"https://openalex.org/S4393918464","display_name":"ArXiv.org","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"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","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4417534296.pdf","grobid_xml":"https://content.openalex.org/works/W4417534296.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"As":[0,108],"organizations":[1],"scale":[2,168],"adoption":[3],"of":[4,49,78,110,117,149,156,219,231],"generative":[5,232],"AI,":[6],"model":[7,119,139],"cost":[8],"optimization":[9],"and":[10,20,104,207,216],"operational":[11],"efficiency":[12],"have":[13],"emerged":[14],"as":[15,99,147],"critical":[16],"factors":[17],"determining":[18],"sustainability":[19],"accessibility.":[21],"While":[22],"Large":[23],"Language":[24,51],"Models":[25,52],"(LLMs)":[26],"demonstrate":[27,160],"impressive":[28],"capabilities":[29],"across":[30],"diverse":[31],"tasks,":[32],"their":[33],"extensive":[34],"computational":[35],"requirements":[36],"make":[37],"them":[38],"cost-prohibitive":[39],"for":[40],"routine":[41],"enterprise":[42],"use.":[43],"This":[44],"limitation":[45],"motivates":[46],"the":[47,76,111,115,124,132,150,165],"exploration":[48],"Small":[50],"(SLMs),":[53],"which":[54],"can":[55,169,221],"deliver":[56],"comparable":[57],"performance":[58,188],"in":[59,145],"targeted":[60,217],"applications":[61],"while":[62],"drastically":[63],"reducing":[64],"infrastructure":[65],"overhead":[66],"(Irugalbandara":[67],"et":[68,176],"al.,":[69,177],"2023).":[70],"In":[71],"this":[72],"work,":[73],"we":[74,113],"investigate":[75],"feasibility":[77],"replacing":[79],"LLM-driven":[80],"workflows":[81],"with":[82,189],"optimized":[83],"SLMs.":[84],"We":[85],"trained":[86],"a":[87,190],"domain-adapted":[88],"SLM":[89,185],"to":[90,172,225],"execute":[91],"representative":[92],"tasks":[93],"traditionally":[94],"handled":[95],"by":[96,142],"LLMs,":[97],"such":[98],"document":[100],"summarization,":[101],"query":[102],"answering,":[103],"structured":[105],"data":[106],"interpretation.":[107],"part":[109,148],"experiment,":[112],"investigated":[114],"fine-tuning":[116],"facebook/opt-350m":[118],"(single":[120],"epoch":[121],"only)":[122],"using":[123],"Hugging":[125],"Face":[126],"TRL":[127],"(Transformer":[128],"Reinforcement":[129],"Learning),":[130],"specifically":[131],"Supervised":[133],"Fine-Tuning":[134],"(SFT)":[135],"trainer.":[136],"The":[137],"OPT-350M":[138],"was":[140],"released":[141],"Meta":[143],"AI":[144,233],"2022":[146],"OPT":[151],"(Open":[152],"Pretrained":[153],"Transformer)":[154],"family":[155],"models.":[157],"Similar":[158],"studies":[159],"that":[161,182,213],"even":[162],"models":[163,201],"at":[164],"350M":[166],"parameter":[167],"meaningfully":[170],"contribute":[171],"instruction-tuning":[173],"pipelines":[174],"(Mekala":[175],"2024).":[178],"Experimental":[179],"results":[180],"demonstrated":[181],"our":[183],"fine-tuned":[184],"achieves":[186],"exceptional":[187],"77.55\\%":[191],"pass":[192],"rate":[193],"on":[194],"ToolBench":[195],"evaluation,":[196],"significantly":[197,222],"outperforming":[198],"all":[199],"baseline":[200],"including":[202],"ChatGPT-CoT":[203],"(26.00\\%),":[204],"ToolLLaMA-DFS":[205],"(30.18\\%),":[206],"ToolLLaMA-CoT":[208],"(16.27\\%).":[209],"These":[210],"findings":[211],"emphasize":[212],"thoughtful":[214],"design":[215],"training":[218],"SLMs":[220],"lower":[223],"barriers":[224],"adoption,":[226],"enabling":[227],"cost-effective,":[228],"large-scale":[229],"integration":[230],"into":[234],"production":[235],"systems.":[236]},"counts_by_year":[],"updated_date":"2026-06-11T09:08:48.828518","created_date":"2025-12-21T00:00:00"}