{"id":"https://openalex.org/W7207797061","doi":"https://doi.org/10.1162/tacl.a.802","title":"Vertical Routing: A Cost-Efficient Collaboration Routing Framework","display_name":"Vertical Routing: A Cost-Efficient Collaboration Routing Framework","publication_year":2026,"publication_date":"2026-01-01","ids":{"openalex":"https://openalex.org/W7207797061","doi":"https://doi.org/10.1162/tacl.a.802"},"language":"en","primary_location":{"id":"doi:10.1162/tacl.a.802","doi":"https://doi.org/10.1162/tacl.a.802","is_oa":true,"landing_page_url":"https://doi.org/10.1162/tacl.a.802","pdf_url":null,"source":{"id":"https://openalex.org/S2729999759","display_name":"Transactions of the Association for Computational Linguistics","issn_l":"2307-387X","issn":["2307-387X"],"is_oa":true,"is_in_doaj":true,"is_core":true,"listed_in":["ccf-b","cwts-core","doaj","dongbi-b","jufo-1","norway-2","poland-20","russia-white-list-1"],"host_organization":"https://openalex.org/P4310320244","host_organization_name":"Association for Computational Linguistics","host_organization_lineage":["https://openalex.org/P4310320244"],"host_organization_lineage_names":["Association for Computational Linguistics"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Transactions of the Association for Computational Linguistics","raw_type":"journal-article"},"volume":null,"type":"article","indexed_in":["crossref","doaj"],"open_access":{"is_oa":true,"oa_status":"diamond","oa_url":"https://doi.org/10.1162/tacl.a.802","any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5149874058","display_name":"Si Shen","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I36399199","display_name":"Nanjing University of Science and Technology","ror":"https://ror.org/00xp9wg62","country_code":"CN","type":"education","lineage":["https://openalex.org/I36399199"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Si Shen","raw_affiliation_strings":["School of Computer Science and Engineering, Nanjing University of Science and Technology, shensi@njust.edu.cn","School of Economics and Management, Nanjing University of Science and Technology"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"School of Computer Science and Engineering, Nanjing University of Science and Technology, shensi@njust.edu.cn","institution_ids":["https://openalex.org/I36399199"]},{"raw_affiliation_string":"School of Economics and Management, Nanjing University of Science and Technology","institution_ids":["https://openalex.org/I36399199"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101297953","display_name":"Peijun Shen","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I36399199","display_name":"Nanjing University of Science and Technology","ror":"https://ror.org/00xp9wg62","country_code":"CN","type":"education","lineage":["https://openalex.org/I36399199"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Peijun Shen","raw_affiliation_strings":["School of Computer Science and Engineering, Nanjing University of Science and Technology, 124106022972@njust.edu.cn"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"School of Computer Science and Engineering, Nanjing University of Science and Technology, 124106022972@njust.edu.cn","institution_ids":["https://openalex.org/I36399199"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5135855855","display_name":"Danhao Zhu","orcid":"https://orcid.org/0009-0003-8642-1593","observed_orcids":["https://orcid.org/0009-0003-8642-1593"]},"institutions":[{"id":"https://openalex.org/I4210145669","display_name":"Shanghai Police College","ror":"https://ror.org/0479fds27","country_code":"CN","type":"education","lineage":["https://openalex.org/I4210145669"]},{"id":"https://openalex.org/I4210157581","display_name":"Jiangsu Police Officer College","ror":"https://ror.org/04k1m2t10","country_code":"CN","type":"education","lineage":["https://openalex.org/I4210157581"]}],"countries":["CN"],"is_corresponding":true,"raw_author_name":"Danhao Zhu","raw_affiliation_strings":["Jiangsu Police Institute, zhudanhao@jspi.cn"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Jiangsu Police Institute, zhudanhao@jspi.cn","institution_ids":["https://openalex.org/I4210145669","https://openalex.org/I4210157581"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":3,"corresponding_author_ids":["https://openalex.org/A5135855855"],"corresponding_institution_ids":["https://openalex.org/I4210145669","https://openalex.org/I4210157581"],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.65495762,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":"14","issue":null,"first_page":"2301","last_page":"2317"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.6240000128746033,"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":0.6240000128746033,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.3278999924659729,"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":0.01810000091791153,"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":[{"id":"https://openalex.org/keywords/small-language-models","display_name":"small language models","score":0.8420000076293945},{"id":"https://openalex.org/keywords/task-decomposition","display_name":"task decomposition","score":0.6449999809265137},{"id":"https://openalex.org/keywords/token-efficiency","display_name":"token efficiency","score":0.6129999756813049},{"id":"https://openalex.org/keywords/generation-cost-reduction","display_name":"generation cost reduction","score":0.550000011920929},{"id":"https://openalex.org/keywords/routing-stability","display_name":"routing stability","score":0.31299999356269836}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8671000003814697},{"id":"https://openalex.org/C48145219","wikidata":"https://www.wikidata.org/wiki/Q1335365","display_name":"Security token","level":2,"score":0.7077999711036682},{"id":"https://openalex.org/C141603448","wikidata":"https://www.wikidata.org/wiki/Q134830","display_name":"Prefix","level":2,"score":0.6984000205993652},{"id":"https://openalex.org/C74172769","wikidata":"https://www.wikidata.org/wiki/Q1446839","display_name":"Routing (electronic design automation)","level":2,"score":0.6704000234603882},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.5940999984741211},{"id":"https://openalex.org/C12725497","wikidata":"https://www.wikidata.org/wiki/Q810247","display_name":"Baseline (sea)","level":2,"score":0.5220999717712402},{"id":"https://openalex.org/C112972136","wikidata":"https://www.wikidata.org/wiki/Q7595718","display_name":"Stability (learning theory)","level":2,"score":0.5080000162124634},{"id":"https://openalex.org/C31258907","wikidata":"https://www.wikidata.org/wiki/Q1301371","display_name":"Computer network","level":1,"score":0.4456000030040741},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.4350999891757965},{"id":"https://openalex.org/C125611927","wikidata":"https://www.wikidata.org/wiki/Q17008131","display_name":"Criticality","level":2,"score":0.39579999446868896},{"id":"https://openalex.org/C204948658","wikidata":"https://www.wikidata.org/wiki/Q1119410","display_name":"Static routing","level":4,"score":0.3677000105381012},{"id":"https://openalex.org/C196423136","wikidata":"https://www.wikidata.org/wiki/Q7209671","display_name":"Policy-based routing","level":5,"score":0.3562999963760376},{"id":"https://openalex.org/C184896649","wikidata":"https://www.wikidata.org/wiki/Q290066","display_name":"Routing table","level":4,"score":0.32260000705718994},{"id":"https://openalex.org/C104954878","wikidata":"https://www.wikidata.org/wiki/Q1648707","display_name":"Routing protocol","level":3,"score":0.32120001316070557},{"id":"https://openalex.org/C2983435990","wikidata":"https://www.wikidata.org/wiki/Q22725","display_name":"Network routing","level":3,"score":0.2985000014305115},{"id":"https://openalex.org/C196083921","wikidata":"https://www.wikidata.org/wiki/Q7915758","display_name":"Variance (accounting)","level":2,"score":0.2912999987602234},{"id":"https://openalex.org/C2984173633","wikidata":"https://www.wikidata.org/wiki/Q22725","display_name":"Routing algorithm","level":4,"score":0.26829999685287476},{"id":"https://openalex.org/C125411270","wikidata":"https://www.wikidata.org/wiki/Q18653","display_name":"Encoding (memory)","level":2,"score":0.26109999418258667},{"id":"https://openalex.org/C2778915421","wikidata":"https://www.wikidata.org/wiki/Q3643177","display_name":"Performance improvement","level":2,"score":0.250900000333786}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1162/tacl.a.802","doi":"https://doi.org/10.1162/tacl.a.802","is_oa":true,"landing_page_url":"https://doi.org/10.1162/tacl.a.802","pdf_url":null,"source":{"id":"https://openalex.org/S2729999759","display_name":"Transactions of the Association for Computational Linguistics","issn_l":"2307-387X","issn":["2307-387X"],"is_oa":true,"is_in_doaj":true,"is_core":true,"listed_in":["ccf-b","cwts-core","doaj","dongbi-b","jufo-1","norway-2","poland-20","russia-white-list-1"],"host_organization":"https://openalex.org/P4310320244","host_organization_name":"Association for Computational Linguistics","host_organization_lineage":["https://openalex.org/P4310320244"],"host_organization_lineage_names":["Association for Computational Linguistics"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Transactions of the Association for Computational Linguistics","raw_type":"journal-article"}],"best_oa_location":{"id":"doi:10.1162/tacl.a.802","doi":"https://doi.org/10.1162/tacl.a.802","is_oa":true,"landing_page_url":"https://doi.org/10.1162/tacl.a.802","pdf_url":null,"source":{"id":"https://openalex.org/S2729999759","display_name":"Transactions of the Association for Computational Linguistics","issn_l":"2307-387X","issn":["2307-387X"],"is_oa":true,"is_in_doaj":true,"is_core":true,"listed_in":["ccf-b","cwts-core","doaj","dongbi-b","jufo-1","norway-2","poland-20","russia-white-list-1"],"host_organization":"https://openalex.org/P4310320244","host_organization_name":"Association for Computational Linguistics","host_organization_lineage":["https://openalex.org/P4310320244"],"host_organization_lineage_names":["Association for Computational Linguistics"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Transactions of the Association for Computational Linguistics","raw_type":"journal-article"},"sustainable_development_goals":[],"sustainable_development_goals_aurora":[{"display_name":"Industry, innovation and infrastructure","score":0.46486490964889526,"id":"https://metadata.un.org/sdg/9"}],"study_designs":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":5,"referenced_works":["https://openalex.org/W3208327127","https://openalex.org/W4221000172","https://openalex.org/W4415428379","https://openalex.org/W4415797314","https://openalex.org/W7126464889"],"related_works":["https://openalex.org/W4416046236","https://openalex.org/W4407185357","https://openalex.org/W7119819136","https://openalex.org/W4416247041","https://openalex.org/W7221126440","https://openalex.org/W7134854408","https://openalex.org/W7092296131","https://openalex.org/W7117870265","https://openalex.org/W4417086258","https://openalex.org/W4407632653"],"abstract_inverted_index":{"Abstract":[0],"Routing":[1],"Large":[2],"and":[3,23,50,77,137,162],"Small":[4],"Language":[5],"Models":[6],"(LLMs":[7],"&":[8],"SLMs)":[9],"is":[10],"commonly":[11],"framed":[12],"as":[13],"query-level":[14],"difficulty":[15,40],"prediction,":[16],"yet":[17],"lightweight":[18],"routers":[19],"are":[20],"often":[21],"unreliable":[22],"stronger":[24],"evaluators":[25],"introduce":[26],"non-trivial":[27],"overhead.":[28],"We":[29],"propose":[30],"VERTICAL":[31,59,116],"ROUTING,":[32],"a":[33,56,69,79,85,92,106],"stage-level":[34],"collaboration":[35],"framework":[36],"that":[37,67,83,115],"avoids":[38],"monolithic":[39],"prediction":[41],"by":[42,123,135,141,147,152],"allocating":[43],"the":[44,52,100,110,119,127,149],"large":[45],"model":[46],"to":[47,55,99,105],"critical":[48,102],"subtasks":[49],"delegating":[51],"remaining":[53],"generation":[54],"small":[57],"model.":[58],"ROUTING":[60,117],"instantiates":[61],"two":[62],"templates:":[63],"(i)":[64],"domain-specific":[65],"templates":[66],"decompose":[68],"task":[70],"into":[71],"ordered":[72],"stages":[73],"with":[74],"criticality":[75],"scores,":[76],"(ii)":[78],"robust":[80],"default":[81,111],"template":[82],"follows":[84],"prefix-first":[86],"prior":[87],"for":[88],"general":[89],"queries.":[90],"Under":[91],"token":[93,133],"budget,":[94],"we":[95],"allocate":[96],"large-model":[97,138],"capacity":[98],"most":[101],"stages,":[103],"or":[104],"budgeted":[107],"prefix":[108],"under":[109],"template.":[112],"Experiments":[113],"show":[114],"outperforms":[118],"strongest":[120],"baseline":[121],"(RouterDC)":[122],"2.0":[124],"points":[125],"in":[126,159],"averaged":[128],"metric,":[129],"while":[130],"significantly":[131],"reducing":[132],"usage":[134],"48.9%":[136],"output":[139],"share":[140],"41.7%.":[142],"Additionally,":[143],"it":[144],"enhances":[145],"stability":[146],"lowering":[148],"standard":[150],"deviation":[151],"87.5%.":[153],"These":[154],"results":[155],"highlight":[156],"its":[157],"advantages":[158],"accuracy,":[160],"efficiency,":[161],"robustness.1":[163]},"counts_by_year":[],"updated_date":"2026-09-24T07:57:37.273135","created_date":"2026-09-04T00:00:00"}