{"id":"https://openalex.org/W7160286340","doi":"https://doi.org/10.48550/arxiv.2605.01897","title":"How Label Imbalance Shapes Geometry: A General Spectral Analysis of Multi-Label Neural Collapse","display_name":"How Label Imbalance Shapes Geometry: A General Spectral Analysis of Multi-Label Neural Collapse","publication_year":2026,"publication_date":"2026-05-03","ids":{"openalex":"https://openalex.org/W7160286340","doi":"https://doi.org/10.48550/arxiv.2605.01897"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2605.01897","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.01897","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2605.01897","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5135386256","display_name":"Xiaoxuan Ma","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ma, Xiaoxuan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5054093367","display_name":"Yixuan Yang","orcid":"https://orcid.org/0000-0002-9334-566X"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yang, Yixuan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135351093","display_name":"Song Li","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Song","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5135305376","display_name":"Xiangyun Hui","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Hui, Xiangyun","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":[],"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/T11550","display_name":"Text and Document Classification Technologies","score":0.6703000068664551,"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/T11550","display_name":"Text and Document Classification Technologies","score":0.6703000068664551,"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/T11652","display_name":"Imbalanced Data Classification Techniques","score":0.04610000178217888,"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/T12535","display_name":"Machine Learning and Data Classification","score":0.03440000116825104,"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/conjecture","display_name":"Conjecture","score":0.5950999855995178},{"id":"https://openalex.org/keywords/covariance","display_name":"Covariance","score":0.5410000085830688},{"id":"https://openalex.org/keywords/stability","display_name":"Stability (learning theory)","score":0.5368000268936157},{"id":"https://openalex.org/keywords/scalar","display_name":"Scalar (mathematics)","score":0.49219998717308044},{"id":"https://openalex.org/keywords/covariance-matrix","display_name":"Covariance matrix","score":0.3716000020503998},{"id":"https://openalex.org/keywords/moment","display_name":"Moment (physics)","score":0.35199999809265137},{"id":"https://openalex.org/keywords/spectrum","display_name":"Spectrum (functional analysis)","score":0.3467000126838684}],"concepts":[{"id":"https://openalex.org/C2780990831","wikidata":"https://www.wikidata.org/wiki/Q319141","display_name":"Conjecture","level":2,"score":0.5950999855995178},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.5449000000953674},{"id":"https://openalex.org/C178650346","wikidata":"https://www.wikidata.org/wiki/Q201984","display_name":"Covariance","level":2,"score":0.5410000085830688},{"id":"https://openalex.org/C112972136","wikidata":"https://www.wikidata.org/wiki/Q7595718","display_name":"Stability (learning theory)","level":2,"score":0.5368000268936157},{"id":"https://openalex.org/C57691317","wikidata":"https://www.wikidata.org/wiki/Q1289248","display_name":"Scalar (mathematics)","level":2,"score":0.49219998717308044},{"id":"https://openalex.org/C185142706","wikidata":"https://www.wikidata.org/wiki/Q1134404","display_name":"Covariance matrix","level":2,"score":0.3716000020503998},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.3637000024318695},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.3596000075340271},{"id":"https://openalex.org/C179254644","wikidata":"https://www.wikidata.org/wiki/Q13222844","display_name":"Moment (physics)","level":2,"score":0.35199999809265137},{"id":"https://openalex.org/C156778621","wikidata":"https://www.wikidata.org/wiki/Q1365748","display_name":"Spectrum (functional analysis)","level":2,"score":0.3467000126838684},{"id":"https://openalex.org/C121864883","wikidata":"https://www.wikidata.org/wiki/Q677916","display_name":"Statistical physics","level":1,"score":0.3449999988079071},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.32850000262260437},{"id":"https://openalex.org/C106487976","wikidata":"https://www.wikidata.org/wiki/Q685816","display_name":"Matrix (chemical analysis)","level":2,"score":0.320499986410141},{"id":"https://openalex.org/C2776502983","wikidata":"https://www.wikidata.org/wiki/Q690182","display_name":"Contrast (vision)","level":2,"score":0.3188999891281128},{"id":"https://openalex.org/C28826006","wikidata":"https://www.wikidata.org/wiki/Q33521","display_name":"Applied mathematics","level":1,"score":0.31349998712539673},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.2851000130176544},{"id":"https://openalex.org/C27438332","wikidata":"https://www.wikidata.org/wiki/Q2873","display_name":"Principal component analysis","level":2,"score":0.2727000117301941},{"id":"https://openalex.org/C18762648","wikidata":"https://www.wikidata.org/wiki/Q42213","display_name":"Work (physics)","level":2,"score":0.2678000032901764},{"id":"https://openalex.org/C152822103","wikidata":"https://www.wikidata.org/wiki/Q7575207","display_name":"Spectral shape analysis","level":3,"score":0.2646999955177307},{"id":"https://openalex.org/C44280652","wikidata":"https://www.wikidata.org/wiki/Q104837","display_name":"Phase (matter)","level":2,"score":0.26170000433921814},{"id":"https://openalex.org/C774472","wikidata":"https://www.wikidata.org/wiki/Q6760393","display_name":"Margin (machine learning)","level":2,"score":0.2558000087738037}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2605.01897","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.01897","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2605.01897","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.01897","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"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":{"This":[0,195],"work":[1,196],"investigates":[2],"the":[3,56,74,108,120,128,134,139,144,151,157,164,173,198,202],"phenomenon":[4],"of":[5,44,111,138,159,201],"Neural":[6,213],"Collapse":[7,214],"(NC)":[8],"in":[9,70],"multi-label":[10,24,36,112,218],"classification,":[11],"extending":[12],"its":[13],"conceptual":[14],"framework":[15,105,210],"from":[16,133],"multi-class":[17],"learning":[18,113],"to":[19,53,106,143,215],"general":[20,115],"correlated":[21],"and":[22,47,64,205],"imbalanced":[23,116,217],"settings.":[25,219],"Although":[26],"recent":[27],"studies":[28],"have":[29],"identified":[30],"a":[31,89,102,125,180,207],"''tag-wise":[32],"averaging''":[33],"structure":[34],"for":[35,55],"features,":[37],"this":[38],"view":[39],"relies":[40],"on":[41,188],"implicit":[42],"assumptions":[43],"label":[45,62,121,140,153],"balance":[46],"combinatorial":[48],"symmetry.":[49],"Consequently,":[50],"it":[51],"fails":[52],"account":[54],"geometrical":[57],"distortions":[58],"caused":[59],"by":[60,79,162],"intrinsic":[61],"correlations":[63],"data":[65],"imbalance,":[66],"which":[67],"are":[68],"common":[69],"practice.":[71],"We":[72,118,170],"resolve":[73],"multiplicity-one":[75],"imbalance":[76,203],"conjecture":[77,204],"raised":[78],"Li":[80],"et":[81],"al.":[82],"(2024),":[83],"showing":[84],"that":[85,150,172,211],"higher-multiplicity":[86],"prototypes":[87],"obey":[88],"class-frequency-weighted":[90],"synthesis":[91],"rule":[92],"rather":[93],"than":[94],"uniform":[95],"averaging.":[96],"To":[97],"address":[98],"this,":[99],"we":[100],"propose":[101],"rigorous":[103],"spectral-control":[104],"analyze":[107],"terminal":[109,160],"phase":[110],"under":[114,183],"conditions.":[117],"introduce":[119],"covariance":[122,154],"spectrum":[123,155],"$\u03ba_m$,":[124],"scalar":[126],"controlling":[127],"distribution-dependent":[129],"lower-bound":[130],"geometry,":[131],"derived":[132],"second-order":[135],"moment":[136],"matrix":[137],"distribution.":[141],"Contrary":[142],"averaging":[145],"perspective,":[146],"our":[147,192],"analysis":[148],"reveals":[149],"centered":[152,166],"controls":[156],"stability":[158],"geometry":[161],"quantifying":[163],"weakest":[165],"inter-class":[167],"contrast":[168],"directions.":[169],"prove":[171],"classical":[174],"Tag-wise":[175],"Averaging":[176],"emerges":[177],"only":[178],"as":[179],"special":[181],"case":[182],"perfect":[184],"orthogonality.":[185],"Numerical":[186],"experiments":[187],"synthetic":[189],"distributions":[190],"validate":[191],"theoretical":[193,209],"bounds.":[194],"resolves":[197],"scaled-average":[199],"aspect":[200],"establishes":[206],"unifying":[208],"extends":[212],"complex,":[216]},"counts_by_year":[],"updated_date":"2026-07-28T07:46:37.118299","created_date":"2026-05-06T00:00:00"}