{"id":"https://openalex.org/W7129091779","doi":"https://doi.org/10.48550/arxiv.2602.12923","title":"Annealing in variational inference mitigates mode collapse: A theoretical study on Gaussian mixtures","display_name":"Annealing in variational inference mitigates mode collapse: A theoretical study on Gaussian mixtures","publication_year":2026,"publication_date":"2026-02-13","ids":{"openalex":"https://openalex.org/W7129091779","doi":"https://doi.org/10.48550/arxiv.2602.12923"},"language":"en","primary_location":{"id":"pmh:doi:10.48550/arxiv.2602.12923","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/A5126092233","display_name":"Luigi Fogliani","orcid":null},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Fogliani, Luigi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5126084550","display_name":"Bruno Loureiro","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Loureiro, Bruno","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5001522727","display_name":"Marylou Gabri\u00e9","orcid":"https://orcid.org/0000-0002-5989-1018"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Gabri\u00e9, Marylou","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":3,"corresponding_author_ids":["https://openalex.org/A5126092233"],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"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/T10775","display_name":"Generative Adversarial Networks and Image Synthesis","score":0.36730000376701355,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"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/T10775","display_name":"Generative Adversarial Networks and Image Synthesis","score":0.36730000376701355,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"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/T12611","display_name":"Neural Networks and Reservoir Computing","score":0.11599999666213989,"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/T12814","display_name":"Gaussian Processes and Bayesian Inference","score":0.10369999706745148,"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/simulated-annealing","display_name":"Simulated annealing","score":0.7559000253677368},{"id":"https://openalex.org/keywords/gaussian","display_name":"Gaussian","score":0.6575000286102295},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.6565999984741211},{"id":"https://openalex.org/keywords/annealing","display_name":"Annealing (glass)","score":0.5625},{"id":"https://openalex.org/keywords/mode","display_name":"Mode (computer interface)","score":0.5220999717712402},{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.4819999933242798}],"concepts":[{"id":"https://openalex.org/C126980161","wikidata":"https://www.wikidata.org/wiki/Q863783","display_name":"Simulated annealing","level":2,"score":0.7559000253677368},{"id":"https://openalex.org/C163716315","wikidata":"https://www.wikidata.org/wiki/Q901177","display_name":"Gaussian","level":2,"score":0.6575000286102295},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.6565999984741211},{"id":"https://openalex.org/C2777855556","wikidata":"https://www.wikidata.org/wiki/Q4339544","display_name":"Annealing (glass)","level":2,"score":0.5625},{"id":"https://openalex.org/C48677424","wikidata":"https://www.wikidata.org/wiki/Q6888088","display_name":"Mode (computer interface)","level":2,"score":0.5220999717712402},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.4819999933242798},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.4586000144481659},{"id":"https://openalex.org/C121864883","wikidata":"https://www.wikidata.org/wiki/Q677916","display_name":"Statistical physics","level":1,"score":0.44429999589920044},{"id":"https://openalex.org/C28826006","wikidata":"https://www.wikidata.org/wiki/Q33521","display_name":"Applied mathematics","level":1,"score":0.4180000126361847},{"id":"https://openalex.org/C61326573","wikidata":"https://www.wikidata.org/wiki/Q1496376","display_name":"Gaussian process","level":3,"score":0.39329999685287476},{"id":"https://openalex.org/C126255220","wikidata":"https://www.wikidata.org/wiki/Q141495","display_name":"Mathematical optimization","level":1,"score":0.3840000033378601},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.3772999942302704},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.30570000410079956},{"id":"https://openalex.org/C2984842247","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep neural networks","level":3,"score":0.30550000071525574},{"id":"https://openalex.org/C66283442","wikidata":"https://www.wikidata.org/wiki/Q1389268","display_name":"Failure mode and effects analysis","level":2,"score":0.2890999913215637},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.2605000138282776},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.25189998745918274}],"mesh":[],"locations_count":3,"locations":[{"id":"pmh:doi:10.48550/arxiv.2602.12923","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":"pmh:oai:HAL:hal-05554700v1","is_oa":true,"landing_page_url":"https://hal.science/hal-05554700","pdf_url":null,"source":{"id":"https://openalex.org/S4406922454","display_name":"SPIRE - Sciences Po Institutional REpository","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":"2026","raw_type":"Preprints, Working Papers, ..."},{"id":"doi:10.48550/arxiv.2602.12923","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2602.12923","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.2602.12923","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":[{"id":"https://openalex.org/G2439744834","display_name":null,"funder_award_id":"ANR-23-IACL-0008","funder_id":"https://openalex.org/F4320320883","funder_display_name":"Agence Nationale de la Recherche"},{"id":"https://openalex.org/G6612625481","display_name":null,"funder_award_id":"France 2030","funder_id":"https://openalex.org/F4320320883","funder_display_name":"Agence Nationale de la Recherche"}],"funders":[{"id":"https://openalex.org/F4320320883","display_name":"Agence Nationale de la Recherche","ror":"https://ror.org/00rbzpz17"},{"id":"https://openalex.org/F4320322892","display_name":"Centre National de la Recherche Scientifique","ror":"https://ror.org/02feahw73"}],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Mode":[0],"collapse,":[1],"the":[2,64,67,71,80],"failure":[3],"to":[4,52,110],"capture":[5],"one":[6],"or":[7],"more":[8],"modes":[9],"when":[10],"targetting":[11],"a":[12,16,28,40,44,55,76],"multimodal":[13],"distribution,":[14],"is":[15,50],"central":[17],"challenge":[18],"in":[19,39,127],"modern":[20],"variational":[21,129],"inference.":[22],"In":[23],"this":[24],"work,":[25],"we":[26,61,100],"provide":[27],"mathematical":[29],"analysis":[30,86],"of":[31,82],"annealing":[32,72,92,122],"based":[33,113],"strategies":[34,123],"for":[35,79,120],"mitigating":[36,124],"mode":[37,48,83,97,125],"collapse":[38,49,126],"tractable":[41],"setting:":[42],"learning":[43],"Gaussian":[45],"mixture,":[46],"where":[47],"known":[51],"arise.":[53],"Leveraging":[54],"low":[56],"dimensional":[57],"summary":[58],"statistics":[59],"description,":[60],"precisely":[62],"characterize":[63],"interplay":[65],"between":[66],"initial":[68],"temperature":[69],"and":[70,74],"rate,":[73],"derive":[75],"sharp":[77],"formula":[78],"probability":[81],"collapse.":[84,98],"Our":[85],"shows":[87],"that":[88,104],"an":[89],"appropriately":[90],"chosen":[91],"scheme":[93],"can":[94],"robustly":[95],"prevent":[96],"Finally,":[99],"present":[101],"numerical":[102],"evidence":[103],"these":[105],"theoretical":[106],"tradeoffs":[107],"qualitatively":[108],"extend":[109],"neural":[111],"network":[112],"models,":[114],"RealNVP":[115],"normalizing":[116],"flows,":[117],"providing":[118],"guidance":[119],"designing":[121],"practical":[128],"inference":[130],"pipelines.":[131]},"counts_by_year":[],"updated_date":"2026-05-01T08:36:08.643496","created_date":"2026-02-17T00:00:00"}