{"id":"https://openalex.org/W7137902888","doi":"https://doi.org/10.1609/aaai.v40i29.39665","title":"How Many Experts Are Enough? Towards Optimal Semantic Specialization for Mixture-of-Experts","display_name":"How Many Experts Are Enough? Towards Optimal Semantic Specialization for Mixture-of-Experts","publication_year":2026,"publication_date":"2026-03-14","ids":{"openalex":"https://openalex.org/W7137902888","doi":"https://doi.org/10.1609/aaai.v40i29.39665"},"language":"en","primary_location":{"id":"doi:10.1609/aaai.v40i29.39665","is_oa":true,"landing_page_url":"https://doi.org/10.1609/aaai.v40i29.39665","pdf_url":null,"source":{"id":"https://openalex.org/S4210191458","display_name":"Proceedings of the AAAI Conference on Artificial Intelligence","issn_l":"2159-5399","issn":["2159-5399","2374-3468"],"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/P4310320058","host_organization_name":"Association for the Advancement of Artificial Intelligence","host_organization_lineage":["https://openalex.org/P4310320058"],"host_organization_lineage_names":["Association for the Advancement of Artificial Intelligence"],"type":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the AAAI Conference on Artificial Intelligence","raw_type":"journal-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"diamond","oa_url":"https://doi.org/10.1609/aaai.v40i29.39665","any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5129731282","display_name":"Sumin Park","orcid":null},"institutions":[{"id":"https://openalex.org/I157485424","display_name":"Korea Advanced Institute of Science and Technology","ror":"https://ror.org/05apxxy63","country_code":"KR","type":"education","lineage":["https://openalex.org/I157485424"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Sumin Park","raw_affiliation_strings":["Korea Advanced Institute of Science & Technology"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Korea Advanced Institute of Science & Technology","institution_ids":["https://openalex.org/I157485424"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5129670180","display_name":"Noseong Park","orcid":null},"institutions":[{"id":"https://openalex.org/I157485424","display_name":"Korea Advanced Institute of Science and Technology","ror":"https://ror.org/05apxxy63","country_code":"KR","type":"education","lineage":["https://openalex.org/I157485424"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Noseong Park","raw_affiliation_strings":["Korea Advanced Institute of Science & Technology"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Korea Advanced Institute of Science & Technology","institution_ids":["https://openalex.org/I157485424"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I157485424"],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.24589416,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":"40","issue":"29","first_page":"24792","last_page":"24800"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11704","display_name":"Mobile Crowdsensing and Crowdsourcing","score":0.5065000057220459,"subfield":{"id":"https://openalex.org/subfields/1706","display_name":"Computer Science Applications"},"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/T11704","display_name":"Mobile Crowdsensing and Crowdsourcing","score":0.5065000057220459,"subfield":{"id":"https://openalex.org/subfields/1706","display_name":"Computer Science Applications"},"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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.2791000008583069,"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/T11273","display_name":"Advanced Graph Neural Networks","score":0.046300001442432404,"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/robustness","display_name":"Robustness (evolution)","score":0.505299985408783},{"id":"https://openalex.org/keywords/subject-matter-expert","display_name":"Subject-matter expert","score":0.4284999966621399},{"id":"https://openalex.org/keywords/flexibility","display_name":"Flexibility (engineering)","score":0.4255000054836273},{"id":"https://openalex.org/keywords/domain","display_name":"Domain (mathematical analysis)","score":0.4246000051498413},{"id":"https://openalex.org/keywords/semantics","display_name":"Semantics (computer science)","score":0.4235999882221222},{"id":"https://openalex.org/keywords/hyperparameter","display_name":"Hyperparameter","score":0.3847000002861023},{"id":"https://openalex.org/keywords/domain-knowledge","display_name":"Domain knowledge","score":0.36719998717308044}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7319999933242798},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5271000266075134},{"id":"https://openalex.org/C63479239","wikidata":"https://www.wikidata.org/wiki/Q7353546","display_name":"Robustness (evolution)","level":3,"score":0.505299985408783},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.47269999980926514},{"id":"https://openalex.org/C105002631","wikidata":"https://www.wikidata.org/wiki/Q4833645","display_name":"Subject-matter expert","level":3,"score":0.4284999966621399},{"id":"https://openalex.org/C2780598303","wikidata":"https://www.wikidata.org/wiki/Q65921492","display_name":"Flexibility (engineering)","level":2,"score":0.4255000054836273},{"id":"https://openalex.org/C36503486","wikidata":"https://www.wikidata.org/wiki/Q11235244","display_name":"Domain (mathematical analysis)","level":2,"score":0.4246000051498413},{"id":"https://openalex.org/C184337299","wikidata":"https://www.wikidata.org/wiki/Q1437428","display_name":"Semantics (computer science)","level":2,"score":0.4235999882221222},{"id":"https://openalex.org/C8642999","wikidata":"https://www.wikidata.org/wiki/Q4171168","display_name":"Hyperparameter","level":2,"score":0.3847000002861023},{"id":"https://openalex.org/C207685749","wikidata":"https://www.wikidata.org/wiki/Q2088941","display_name":"Domain knowledge","level":2,"score":0.36719998717308044},{"id":"https://openalex.org/C58328972","wikidata":"https://www.wikidata.org/wiki/Q184609","display_name":"Expert system","level":2,"score":0.3434000015258789},{"id":"https://openalex.org/C204323151","wikidata":"https://www.wikidata.org/wiki/Q905424","display_name":"Range (aeronautics)","level":2,"score":0.33000001311302185},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.321399986743927},{"id":"https://openalex.org/C94124525","wikidata":"https://www.wikidata.org/wiki/Q912550","display_name":"Categorization","level":2,"score":0.31029999256134033},{"id":"https://openalex.org/C2778180026","wikidata":"https://www.wikidata.org/wiki/Q18378163","display_name":"Semantic heterogeneity","level":4,"score":0.30410000681877136},{"id":"https://openalex.org/C28719098","wikidata":"https://www.wikidata.org/wiki/Q44946","display_name":"Point (geometry)","level":2,"score":0.2980000078678131},{"id":"https://openalex.org/C2522767166","wikidata":"https://www.wikidata.org/wiki/Q2374463","display_name":"Data science","level":1,"score":0.2822999954223633},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.28049999475479126},{"id":"https://openalex.org/C23123220","wikidata":"https://www.wikidata.org/wiki/Q816826","display_name":"Information retrieval","level":1,"score":0.2646999955177307},{"id":"https://openalex.org/C511149849","wikidata":"https://www.wikidata.org/wiki/Q7449051","display_name":"Semantic computing","level":3,"score":0.2556999921798706}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.1609/aaai.v40i29.39665","is_oa":true,"landing_page_url":"https://doi.org/10.1609/aaai.v40i29.39665","pdf_url":null,"source":{"id":"https://openalex.org/S4210191458","display_name":"Proceedings of the AAAI Conference on Artificial Intelligence","issn_l":"2159-5399","issn":["2159-5399","2374-3468"],"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/P4310320058","host_organization_name":"Association for the Advancement of Artificial Intelligence","host_organization_lineage":["https://openalex.org/P4310320058"],"host_organization_lineage_names":["Association for the Advancement of Artificial Intelligence"],"type":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the AAAI Conference on Artificial Intelligence","raw_type":"journal-article"},{"id":"pmh:oai:ojs.aaai.org:article/39665","is_oa":false,"landing_page_url":"https://ojs.aaai.org/index.php/AAAI/article/view/39665","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"2159-5399","raw_type":"info:eu-repo/semantics/article"}],"best_oa_location":{"id":"doi:10.1609/aaai.v40i29.39665","is_oa":true,"landing_page_url":"https://doi.org/10.1609/aaai.v40i29.39665","pdf_url":null,"source":{"id":"https://openalex.org/S4210191458","display_name":"Proceedings of the AAAI Conference on Artificial Intelligence","issn_l":"2159-5399","issn":["2159-5399","2374-3468"],"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/P4310320058","host_organization_name":"Association for the Advancement of Artificial Intelligence","host_organization_lineage":["https://openalex.org/P4310320058"],"host_organization_lineage_names":["Association for the Advancement of Artificial Intelligence"],"type":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the AAAI Conference on Artificial Intelligence","raw_type":"journal-article"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Finding":[0],"the":[1,19,37,47,87,95,100,129],"optimal":[2,132],"configuration":[3],"of":[4,22,39,99,106,131,166],"Sparse":[5],"Mixture-of-":[6],"Experts":[7],"(SMoE)":[8],"that":[9,81,110,124,160],"maximizes":[10],"semantic":[11,41,78,97],"differentiation":[12],"among":[13],"experts":[14,44],"is":[15],"essential":[16],"for":[17,54,63],"exploiting":[18],"full":[20,96],"potential":[21],"MoE":[23,61,168],"architectures.":[24],"However,":[25],"existing":[26,88],"SMoE":[27],"frameworks":[28],"either":[29],"heavily":[30],"rely":[31],"on":[32,116,151],"hyperparameter":[33],"tuning":[34],"or":[35],"overlook":[36],"importance":[38],"diversifying":[40],"roles":[42],"across":[43,154],"when":[45,86],"adapting":[46],"expert":[48,65,84,89,113,176],"pool":[49,90],"size.":[50],"We":[51,121],"propose":[52],"Mixture-of-Experts":[53],"Adaptive":[55],"Semantic":[56],"Specialization":[57],"(MASS),":[58],"a":[59,76,143,164],"semantic-aware":[60],"framework":[62],"adaptive":[64,107],"expansion":[66,85],"and":[67,102,156,174],"dynamic":[68],"routing.":[69],"MASS":[70,125,161],"introduces":[71],"two":[72],"key":[73],"advancements:":[74],"(i)":[75],"gradient-based":[77],"drift":[79],"detector":[80],"prompts":[82],"targeted":[83],"lacks":[91],"capacity":[92],"to":[93,128],"capture":[94],"diversity":[98],"data,":[101],"(ii)":[103],"an":[104],"integration":[105],"routing":[108,118],"strategy":[109],"dynamically":[111],"adjusts":[112],"usage":[114],"based":[115],"token-level":[117],"confidence":[119],"mass.":[120],"first":[122],"demonstrate":[123],"reliably":[126],"converges":[127],"point":[130],"balance":[133],"between":[134],"cost-performance":[135],"trade-off":[136],"with":[137],"notably":[138],"improved":[139],"sematic":[140],"specialization":[141],"in":[142],"highly":[144],"controlled":[145],"synthetic":[146],"setup.":[147],"Further":[148],"empirical":[149],"results":[150],"real-world":[152],"datasets":[153],"language":[155],"vision":[157],"domains":[158],"show":[159],"consistently":[162],"outperforms":[163],"range":[165],"strong":[167],"baselines,":[169],"demonstrating":[170],"its":[171],"domain":[172],"robustness":[173],"enhanced":[175],"specialization.":[177]},"counts_by_year":[],"updated_date":"2026-07-29T14:22:42.915294","created_date":"2026-03-18T00:00:00"}