{"id":"https://openalex.org/W7225822208","doi":"https://doi.org/10.5281/zenodo.24137263","title":"A Pooled Accuracy Is Not a Scaling Verdict: What Tool-Calling Benchmarks Show About Whether Larger Models Call Tools Better, Why Per-Token Price Misranks Cost Only in One Cost Regime, and the Per-Skill, Per-Task Accounting a Router Needs","display_name":"A Pooled Accuracy Is Not a Scaling Verdict: What Tool-Calling Benchmarks Show About Whether Larger Models Call Tools Better, Why Per-Token Price Misranks Cost Only in One Cost Regime, and the Per-Skill, Per-Task Accounting a Router Needs","publication_year":2026,"publication_date":"2026-11-11","ids":{"openalex":"https://openalex.org/W7225822208","doi":"https://doi.org/10.5281/zenodo.24137263"},"language":"en","primary_location":{"id":"doi:10.5281/zenodo.24137263","doi":"https://doi.org/10.5281/zenodo.24137263","is_oa":true,"landing_page_url":"https://doi.org/10.5281/zenodo.24137263","pdf_url":null,"source":{"id":"https://openalex.org/S4306400562","display_name":"Zenodo (CERN European Organization for Nuclear Research)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I67311998","host_organization_name":"European Organization for Nuclear Research","host_organization_lineage":["https://openalex.org/I67311998"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"volume":null,"type":"preprint","is_traditional":true,"indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.5281/zenodo.24137263","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5147720734","display_name":"Pranay M. Mahendrakar","orcid":"https://orcid.org/0009-0003-7224-029X","observed_orcids":["https://orcid.org/0009-0003-7224-029X"]},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Mahendrakar, Pranay","raw_affiliation_strings":["SONYTECH"],"raw_orcid":"https://orcid.org/0009-0003-7224-029X","affiliations":[{"raw_affiliation_string":"SONYTECH","institution_ids":[]}]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":["https://openalex.org/A5147720734"],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.3463999927043915,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.3463999927043915,"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/T10028","display_name":"Topic Modeling","score":0.33889999985694885,"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/T10286","display_name":"Information Retrieval and Search Behavior","score":0.06430000066757202,"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"}}],"keywords":[],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6193000078201294},{"id":"https://openalex.org/C149782125","wikidata":"https://www.wikidata.org/wiki/Q160039","display_name":"Econometrics","level":1,"score":0.47450000047683716},{"id":"https://openalex.org/C198082294","wikidata":"https://www.wikidata.org/wiki/Q3399648","display_name":"Position (finance)","level":2,"score":0.38960000872612},{"id":"https://openalex.org/C99844830","wikidata":"https://www.wikidata.org/wiki/Q102441924","display_name":"Scaling","level":2,"score":0.3693000078201294},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.3361999988555908},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.31360000371932983},{"id":"https://openalex.org/C140779682","wikidata":"https://www.wikidata.org/wiki/Q210868","display_name":"Sampling (signal processing)","level":3,"score":0.31209999322891235},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.31130000948905945},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3073999881744385},{"id":"https://openalex.org/C2522767166","wikidata":"https://www.wikidata.org/wiki/Q2374463","display_name":"Data science","level":1,"score":0.30489999055862427},{"id":"https://openalex.org/C105795698","wikidata":"https://www.wikidata.org/wiki/Q12483","display_name":"Statistics","level":1,"score":0.2962999939918518},{"id":"https://openalex.org/C133462117","wikidata":"https://www.wikidata.org/wiki/Q4929239","display_name":"Data collection","level":2,"score":0.28279998898506165},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.28119999170303345},{"id":"https://openalex.org/C162118730","wikidata":"https://www.wikidata.org/wiki/Q1128453","display_name":"Actuarial science","level":1,"score":0.27300000190734863},{"id":"https://openalex.org/C2779530757","wikidata":"https://www.wikidata.org/wiki/Q1207505","display_name":"Quality (philosophy)","level":2,"score":0.2705000042915344},{"id":"https://openalex.org/C91682802","wikidata":"https://www.wikidata.org/wiki/Q620538","display_name":"Multidimensional scaling","level":2,"score":0.2671000063419342},{"id":"https://openalex.org/C1893757","wikidata":"https://www.wikidata.org/wiki/Q3653001","display_name":"Inversion (geology)","level":3,"score":0.25839999318122864},{"id":"https://openalex.org/C2775896111","wikidata":"https://www.wikidata.org/wiki/Q642560","display_name":"Router","level":2,"score":0.2567000091075897}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.5281/zenodo.24137263","doi":"https://doi.org/10.5281/zenodo.24137263","is_oa":true,"landing_page_url":"https://doi.org/10.5281/zenodo.24137263","pdf_url":null,"source":{"id":"https://openalex.org/S4306400562","display_name":"Zenodo (CERN European Organization for Nuclear Research)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I67311998","host_organization_name":"European Organization for Nuclear Research","host_organization_lineage":["https://openalex.org/I67311998"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},{"id":"doi:10.5281/zenodo.24137264","doi":"https://doi.org/10.5281/zenodo.24137264","is_oa":true,"landing_page_url":"https://doi.org/10.5281/zenodo.24137264","pdf_url":null,"source":{"id":"https://openalex.org/S4306400562","display_name":"Zenodo (CERN European Organization for Nuclear Research)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I67311998","host_organization_name":"European Organization for Nuclear Research","host_organization_lineage":["https://openalex.org/I67311998"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.5281/zenodo.24137263","doi":"https://doi.org/10.5281/zenodo.24137263","is_oa":true,"landing_page_url":"https://doi.org/10.5281/zenodo.24137263","pdf_url":null,"source":{"id":"https://openalex.org/S4306400562","display_name":"Zenodo (CERN European Organization for Nuclear Research)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I67311998","host_organization_name":"European Organization for Nuclear Research","host_organization_lineage":["https://openalex.org/I67311998"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"(c)":[0],"2026":[1,64],"Pranay":[2],"Mahendrakar.":[3],"Licensed":[4],"under":[5,115],"CC":[6],"BY":[7],"4.0.":[8],"Life":[9],"of":[10,94,132,146,161,179,184,282,294],"Research":[11],"Two":[12],"claims":[13,111],"about":[14,129,176],"tool-using":[15],"language":[16],"models":[17,29,59,74,90],"now":[18],"circulate":[19],"together.":[20],"The":[21,41,189,296],"first":[22,127],"is":[23,43,47,175,195],"that":[24,44,88,109,117,174,192,335],"larger":[25],"or":[26,38],"more":[27,241],"expensive":[28],"do":[30],"not":[31,227,233],"reliably":[32,235],"call":[33,216,229],"tools":[34],"better":[35],"than":[36],"smaller":[37],"cheaper":[39],"ones.":[40],"second":[42,259],"per-token":[45],"price":[46,268],"a":[48,53,81,172,222,230,280,324],"poor":[49],"guide":[50],"to":[51,107,228,284],"what":[52],"request":[54],"costs,":[55],"because":[56],"some":[57],"cheap":[58,148],"spend":[60,185],"many":[61],"tokens.":[62],"A":[63],"router":[65],"for":[66,121,186,269],"agentic":[67,95],"tool":[68,231,250],"calling":[69],"reported":[70],"both,":[71],"from":[72,141],"eight":[73],"scored":[75],"on":[76,85,154,165,191,247],"five":[77],"function-calling":[78],"benchmarks,":[79],"and":[80,103,181,238,245,273,305,331],"position":[82],"paper":[83],"argues":[84],"related":[86],"grounds":[87],"small":[89,193],"are":[91,112],"the":[92,100,118,126,133,136,151,155,166,200,207,210,225,258,261,299,309,317,333,338],"future":[93],"systems.":[96],"In":[97],"this":[98],"analysis":[99],"author":[101,297],"reviews":[102],"synthesises":[104],"published":[105,311],"results":[106,312],"show":[108],"both":[110,160],"true":[113],"only":[114],"conditions":[116],"evidence":[119],"offered":[120],"them":[122],"rarely":[123],"states.":[124],"On":[125,257],"claim,":[128,260],"90":[130],"percent":[131,178],"items":[134],"behind":[135],"router's":[137,201,262,310],"pooled":[138,211],"accuracy":[139,217],"come":[140],"two":[142,292],"public":[143],"fine-tuning":[144],"corpora":[145],"short,":[147],"calls;":[149],"by":[150,279,291],"author's":[152],"arithmetic":[153],"study's":[156],"own":[157,263],"per-group":[158],"table,":[159],"its":[162,270],"inversions":[163],"reverse":[164],"multi-turn":[167],"subset":[168,173,194],"it":[169,290],"calls":[170],"hardest,":[171],"1":[177,304],"queries":[180],"roughly":[182],"three-quarters":[183],"most":[187],"models.":[188],"reversal":[190],"within":[196,221],"sampling":[197],"noise,":[198],"so":[199],"data":[202],"neither":[203],"establish":[204],"nor":[205],"refute":[206],"inversion":[208],"outside":[209],"mix.":[212],"Across":[213],"other":[214],"studies,":[215],"rises":[218],"with":[219,236,242],"size":[220,237],"model":[223,254],"family;":[224],"decision":[226],"does":[232],"improve":[234],"moves":[239,276,289],"far":[240],"targeted":[243],"training;":[244],"switching":[246],"reasoning":[248],"raised":[249],"hallucination":[251],"in":[252,302,313,320],"every":[253],"pair":[255],"tested.":[256],"cost":[264,278],"ordering":[265],"matches":[266],"list":[267],"main":[271],"pool,":[272],"output":[274],"verbosity":[275],"per-request":[277],"factor":[281],"0.78":[283],"1.31,":[285],"while":[286],"conversation":[287],"length":[288],"orders":[293],"magnitude.":[295],"consolidates":[298],"measured":[300],"values":[301],"Table":[303,314,321],"Figure":[306],"1,":[307,330],"re-weights":[308],"2,":[315],"separates":[316],"tool-calling":[318],"sub-skills":[319],"3,":[322],"states":[323],"per-skill,":[325],"per-task":[326],"accounting":[327],"as":[328],"Algorithm":[329],"names":[332],"measurements":[334],"would":[336],"settle":[337],"open":[339],"part.":[340]},"counts_by_year":[],"updated_date":"2026-11-12T12:28:57.059296","created_date":"2026-11-12T00:00:00"}