{"id":"https://openalex.org/W7127602015","doi":"https://doi.org/10.48550/arxiv.2602.03204","title":"Sparsity is Combinatorial Depth: Quantifying MoE Expressivity via Tropical Geometry","display_name":"Sparsity is Combinatorial Depth: Quantifying MoE Expressivity via Tropical Geometry","publication_year":2026,"publication_date":"2026-02-03","ids":{"openalex":"https://openalex.org/W7127602015","doi":"https://doi.org/10.48550/arxiv.2602.03204"},"language":null,"primary_location":{"id":"pmh:doi:10.48550/arxiv.2602.03204","is_oa":true,"landing_page_url":null,"pdf_url":null,"source":{"id":"https://openalex.org/S4406922384","display_name":"Open MIND","issn_l":null,"issn":null,"is_oa":false,"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":"Article"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":null,"any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5125040046","display_name":"Ye Su","orcid":null},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Su, Ye","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5125026274","display_name":"Huayi Tang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Tang, Huayi","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5125022391","display_name":"Zixuan Gong","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Gong, Zixuan","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5125001309","display_name":"Yong Liu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liu, Yong","raw_affiliation_strings":[],"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":4,"corresponding_author_ids":["https://openalex.org/A5125040046"],"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/T12808","display_name":"Ferroelectric and Negative Capacitance Devices","score":0.1859000027179718,"subfield":{"id":"https://openalex.org/subfields/2208","display_name":"Electrical and Electronic Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T12808","display_name":"Ferroelectric and Negative Capacitance Devices","score":0.1859000027179718,"subfield":{"id":"https://openalex.org/subfields/2208","display_name":"Electrical and Electronic Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T11435","display_name":"Polynomial and algebraic computation","score":0.13740000128746033,"subfield":{"id":"https://openalex.org/subfields/1703","display_name":"Computational Theory and Mathematics"},"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/T12536","display_name":"Topological and Geometric Data Analysis","score":0.1354999989271164,"subfield":{"id":"https://openalex.org/subfields/1703","display_name":"Computational Theory and Mathematics"},"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/tropical-geometry","display_name":"Tropical geometry","score":0.5299000144004822},{"id":"https://openalex.org/keywords/transversality","display_name":"Transversality","score":0.5289000272750854},{"id":"https://openalex.org/keywords/routing","display_name":"Routing (electronic design automation)","score":0.44940000772476196},{"id":"https://openalex.org/keywords/heuristic","display_name":"Heuristic","score":0.4185999929904938},{"id":"https://openalex.org/keywords/manifold","display_name":"Manifold (fluid mechanics)","score":0.3935999870300293},{"id":"https://openalex.org/keywords/topology","display_name":"Topology (electrical circuits)","score":0.3889000117778778},{"id":"https://openalex.org/keywords/isomorphism","display_name":"Isomorphism (crystallography)","score":0.38370001316070557},{"id":"https://openalex.org/keywords/space","display_name":"Space (punctuation)","score":0.3391999900341034}],"concepts":[{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.6851000189781189},{"id":"https://openalex.org/C202652594","wikidata":"https://www.wikidata.org/wiki/Q2982669","display_name":"Tropical geometry","level":2,"score":0.5299000144004822},{"id":"https://openalex.org/C24167531","wikidata":"https://www.wikidata.org/wiki/Q2300767","display_name":"Transversality","level":2,"score":0.5289000272750854},{"id":"https://openalex.org/C74172769","wikidata":"https://www.wikidata.org/wiki/Q1446839","display_name":"Routing (electronic design automation)","level":2,"score":0.44940000772476196},{"id":"https://openalex.org/C173801870","wikidata":"https://www.wikidata.org/wiki/Q201413","display_name":"Heuristic","level":2,"score":0.4185999929904938},{"id":"https://openalex.org/C529865628","wikidata":"https://www.wikidata.org/wiki/Q1790740","display_name":"Manifold (fluid mechanics)","level":2,"score":0.3935999870300293},{"id":"https://openalex.org/C184720557","wikidata":"https://www.wikidata.org/wiki/Q7825049","display_name":"Topology (electrical circuits)","level":2,"score":0.3889000117778778},{"id":"https://openalex.org/C203436722","wikidata":"https://www.wikidata.org/wiki/Q902950","display_name":"Isomorphism (crystallography)","level":3,"score":0.38370001316070557},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.34299999475479126},{"id":"https://openalex.org/C2778572836","wikidata":"https://www.wikidata.org/wiki/Q380933","display_name":"Space (punctuation)","level":2,"score":0.3391999900341034},{"id":"https://openalex.org/C2776477805","wikidata":"https://www.wikidata.org/wiki/Q4460773","display_name":"Topological data analysis","level":2,"score":0.3319000005722046},{"id":"https://openalex.org/C182950735","wikidata":"https://www.wikidata.org/wiki/Q1930978","display_name":"Tensegrity","level":2,"score":0.329800009727478},{"id":"https://openalex.org/C2781315470","wikidata":"https://www.wikidata.org/wiki/Q193623","display_name":"Binomial (polynomial)","level":2,"score":0.31130000948905945},{"id":"https://openalex.org/C81332173","wikidata":"https://www.wikidata.org/wiki/Q179899","display_name":"Topological space","level":2,"score":0.3111000061035156},{"id":"https://openalex.org/C148043351","wikidata":"https://www.wikidata.org/wiki/Q4456944","display_name":"Current (fluid)","level":2,"score":0.3043000102043152},{"id":"https://openalex.org/C68363185","wikidata":"https://www.wikidata.org/wiki/Q180969","display_name":"Algebraic geometry","level":2,"score":0.2957000136375427},{"id":"https://openalex.org/C2524010","wikidata":"https://www.wikidata.org/wiki/Q8087","display_name":"Geometry","level":1,"score":0.29409998655319214},{"id":"https://openalex.org/C126255220","wikidata":"https://www.wikidata.org/wiki/Q141495","display_name":"Mathematical optimization","level":1,"score":0.2930999994277954},{"id":"https://openalex.org/C118615104","wikidata":"https://www.wikidata.org/wiki/Q121416","display_name":"Discrete mathematics","level":1,"score":0.2921000123023987},{"id":"https://openalex.org/C158843486","wikidata":"https://www.wikidata.org/wiki/Q2137810","display_name":"Complex geometry","level":2,"score":0.2711000144481659},{"id":"https://openalex.org/C193319292","wikidata":"https://www.wikidata.org/wiki/Q272172","display_name":"Hamming distance","level":2,"score":0.267300009727478},{"id":"https://openalex.org/C156135169","wikidata":"https://www.wikidata.org/wiki/Q5535522","display_name":"Geometry and topology","level":2,"score":0.266400009393692},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.26109999418258667},{"id":"https://openalex.org/C13336665","wikidata":"https://www.wikidata.org/wiki/Q125977","display_name":"Vector space","level":2,"score":0.26019999384880066},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.25929999351501465}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:doi:10.48550/arxiv.2602.03204","is_oa":true,"landing_page_url":null,"pdf_url":null,"source":{"id":"https://openalex.org/S4406922384","display_name":"Open MIND","issn_l":null,"issn":null,"is_oa":false,"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":"Article"},{"id":"doi:10.48550/arxiv.2602.03204","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2602.03204","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:doi:10.48550/arxiv.2602.03204","is_oa":true,"landing_page_url":null,"pdf_url":null,"source":{"id":"https://openalex.org/S4406922384","display_name":"Open MIND","issn_l":null,"issn":null,"is_oa":false,"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":"Article"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"While":[0],"Mixture-of-Experts":[1],"(MoE)":[2],"architectures":[3,109],"define":[4],"the":[5,25,31,38,46,55,59,76,86,92,117,128,132,135,147],"state-of-the-art,":[6],"their":[7],"theoretical":[8,144],"success":[9],"is":[10,42,68],"often":[11],"attributed":[12],"to":[13,45],"heuristic":[14],"efficiency":[15],"rather":[16],"than":[17],"geometric":[18,73],"expressivity.":[19],"In":[20,122],"this":[21,123],"work,":[22],"we":[23,84],"present":[24],"first":[26],"analysis":[27],"of":[28,33,62,88,119,131,138,150],"MoE":[29,108],"through":[30],"lens":[32],"tropical":[34,50],"geometry,":[35],"establishing":[36],"that":[37,66,97],"Top-$k$":[39],"routing":[40,120],"mechanism":[41],"algebraically":[43],"isomorphic":[44],"$k$-th":[47],"elementary":[48],"symmetric":[49],"polynomial.":[51],"This":[52],"isomorphism":[53],"partitions":[54],"input":[56],"space":[57],"into":[58],"Normal":[60],"Fan":[61],"a":[63,142],"Hypersimplex,":[64],"revealing":[65],"\\textbf{sparsity":[67],"combinatorial":[69],"depth}":[70],"which":[71],"scales":[72],"capacity":[74,103],"by":[75],"binomial":[77],"coefficient":[78],"$\\binom{N}{k}$.":[79],"Moving":[80],"beyond":[81],"ambient":[82],"bounds,":[83],"introduce":[85],"concept":[87],"\\textit{Effective":[89],"Capacity}":[90],"under":[91],"Manifold":[93],"Hypothesis.":[94],"We":[95],"prove":[96],"while":[98],"dense":[99],"networks":[100],"suffer":[101],"from":[102],"collapse":[104],"on":[105],"low-dimensional":[106],"data,":[107],"exhibit":[110],"\\textit{Combinatorial":[111],"Resilience},":[112],"maintaining":[113],"high":[114],"expressivity":[115],"via":[116],"transversality":[118],"cones.":[121],"study,":[124],"our":[125],"framework":[126],"unifies":[127],"discrete":[129],"geometry":[130,137],"Hypersimplex":[133],"with":[134],"continuous":[136],"neural":[139],"functions,":[140],"offering":[141],"rigorous":[143],"justification":[145],"for":[146],"topological":[148],"supremacy":[149],"conditional":[151],"computation.":[152]},"counts_by_year":[],"updated_date":"2026-04-04T16:13:02.066488","created_date":"2026-02-06T00:00:00"}
