{"id":"https://openalex.org/W7161829937","doi":"https://doi.org/10.48550/arxiv.2605.18931","title":"Markov Chain Decoders Overcome the Heavy-Tail Limitations of Lipschitz Generative Models","display_name":"Markov Chain Decoders Overcome the Heavy-Tail Limitations of Lipschitz Generative Models","publication_year":2026,"publication_date":"2026-05-18","ids":{"openalex":"https://openalex.org/W7161829937","doi":"https://doi.org/10.48550/arxiv.2605.18931"},"language":"en","primary_location":{"id":"pmh:oai:HAL:hal-05620472v1","is_oa":false,"landing_page_url":"https://hal.science/hal-05620472","pdf_url":null,"source":{"id":"https://openalex.org/S4306402512","display_name":"HAL (Le Centre pour la Communication Scientifique Directe)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I1294671590","host_organization_name":"Centre National de la Recherche Scientifique","host_organization_lineage":["https://openalex.org/I1294671590"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"22nd European Performance Engineering Workshop (EPEW 2026), Jun 2025, Grimstad, Norway","raw_type":"info:eu-repo/semantics/conferenceObject"},"type":"conference-paper","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2605.18931","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5072587293","display_name":"A. Ziani","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ziani, Abdelhakim","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5055547076","display_name":"Andr\u00e1s Horv\u00e1th","orcid":"https://orcid.org/0000-0002-8210-545X"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Horvath, Andras","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5067435617","display_name":"Paolo Ballarini","orcid":"https://orcid.org/0000-0002-4652-3953"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ballarini, Paolo","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":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.61619677,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"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":0.6894000172615051,"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":0.6894000172615051,"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/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.060100000351667404,"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/T11612","display_name":"Stochastic Gradient Optimization Techniques","score":0.032600000500679016,"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/lipschitz-continuity","display_name":"Lipschitz continuity","score":0.732699990272522},{"id":"https://openalex.org/keywords/gaussian","display_name":"Gaussian","score":0.6776000261306763},{"id":"https://openalex.org/keywords/markov-chain","display_name":"Markov chain","score":0.6395999789237976},{"id":"https://openalex.org/keywords/generative-model","display_name":"Generative model","score":0.44769999384880066},{"id":"https://openalex.org/keywords/gaussian-process","display_name":"Gaussian process","score":0.397599995136261},{"id":"https://openalex.org/keywords/hidden-markov-model","display_name":"Hidden Markov model","score":0.3741999864578247},{"id":"https://openalex.org/keywords/latent-variable","display_name":"Latent variable","score":0.3707999885082245},{"id":"https://openalex.org/keywords/quantile","display_name":"Quantile","score":0.3400999903678894},{"id":"https://openalex.org/keywords/markov-model","display_name":"Markov model","score":0.3337000012397766}],"concepts":[{"id":"https://openalex.org/C22324862","wikidata":"https://www.wikidata.org/wiki/Q652707","display_name":"Lipschitz continuity","level":2,"score":0.732699990272522},{"id":"https://openalex.org/C163716315","wikidata":"https://www.wikidata.org/wiki/Q901177","display_name":"Gaussian","level":2,"score":0.6776000261306763},{"id":"https://openalex.org/C98763669","wikidata":"https://www.wikidata.org/wiki/Q176645","display_name":"Markov chain","level":2,"score":0.6395999789237976},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5558000206947327},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.53329998254776},{"id":"https://openalex.org/C167966045","wikidata":"https://www.wikidata.org/wiki/Q5532625","display_name":"Generative model","level":3,"score":0.44769999384880066},{"id":"https://openalex.org/C61326573","wikidata":"https://www.wikidata.org/wiki/Q1496376","display_name":"Gaussian process","level":3,"score":0.397599995136261},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.38029998540878296},{"id":"https://openalex.org/C23224414","wikidata":"https://www.wikidata.org/wiki/Q176769","display_name":"Hidden Markov model","level":2,"score":0.3741999864578247},{"id":"https://openalex.org/C51167844","wikidata":"https://www.wikidata.org/wiki/Q4422623","display_name":"Latent variable","level":2,"score":0.3707999885082245},{"id":"https://openalex.org/C118671147","wikidata":"https://www.wikidata.org/wiki/Q578714","display_name":"Quantile","level":2,"score":0.3400999903678894},{"id":"https://openalex.org/C126255220","wikidata":"https://www.wikidata.org/wiki/Q141495","display_name":"Mathematical optimization","level":1,"score":0.3353999853134155},{"id":"https://openalex.org/C163836022","wikidata":"https://www.wikidata.org/wiki/Q6771326","display_name":"Markov model","level":3,"score":0.3337000012397766},{"id":"https://openalex.org/C57273362","wikidata":"https://www.wikidata.org/wiki/Q576722","display_name":"Decoding methods","level":2,"score":0.3325999975204468},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.33180001378059387},{"id":"https://openalex.org/C8642999","wikidata":"https://www.wikidata.org/wiki/Q4171168","display_name":"Hyperparameter","level":2,"score":0.32519999146461487},{"id":"https://openalex.org/C110121322","wikidata":"https://www.wikidata.org/wiki/Q865811","display_name":"Distribution (mathematics)","level":2,"score":0.31709998846054077},{"id":"https://openalex.org/C159886148","wikidata":"https://www.wikidata.org/wiki/Q176645","display_name":"Markov process","level":2,"score":0.30320000648498535},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3027999997138977},{"id":"https://openalex.org/C187691185","wikidata":"https://www.wikidata.org/wiki/Q2020720","display_name":"Grid","level":2,"score":0.2906000018119812},{"id":"https://openalex.org/C61224824","wikidata":"https://www.wikidata.org/wiki/Q2260434","display_name":"Mixture model","level":2,"score":0.2888000011444092},{"id":"https://openalex.org/C189973286","wikidata":"https://www.wikidata.org/wiki/Q176695","display_name":"Markov property","level":4,"score":0.2782999873161316},{"id":"https://openalex.org/C39890363","wikidata":"https://www.wikidata.org/wiki/Q36108","display_name":"Generative grammar","level":2,"score":0.27799999713897705},{"id":"https://openalex.org/C54907487","wikidata":"https://www.wikidata.org/wiki/Q7915688","display_name":"Variable-order Markov model","level":4,"score":0.27399998903274536},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.2700999975204468},{"id":"https://openalex.org/C149441793","wikidata":"https://www.wikidata.org/wiki/Q200726","display_name":"Probability distribution","level":2,"score":0.2612999975681305},{"id":"https://openalex.org/C28826006","wikidata":"https://www.wikidata.org/wiki/Q33521","display_name":"Applied mathematics","level":1,"score":0.2574999928474426}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:HAL:hal-05620472v1","is_oa":false,"landing_page_url":"https://hal.science/hal-05620472","pdf_url":null,"source":{"id":"https://openalex.org/S4306402512","display_name":"HAL (Le Centre pour la Communication Scientifique Directe)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I1294671590","host_organization_name":"Centre National de la Recherche Scientifique","host_organization_lineage":["https://openalex.org/I1294671590"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"22nd European Performance Engineering Workshop (EPEW 2026), Jun 2025, Grimstad, Norway","raw_type":"info:eu-repo/semantics/conferenceObject"},{"id":"doi:10.48550/arxiv.2605.18931","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.18931","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.2605.18931","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.18931","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":{"Heavy-tailed":[0],"distributions":[1,135,185],"are":[2],"prevalent":[3],"in":[4],"performance":[5],"evaluation,":[6],"network":[7],"traffic,":[8],"and":[9,31,50,79,100,130,162,196],"risk":[10],"modeling.":[11],"This":[12],"behavior":[13],"poses":[14],"a":[15,35,73,80,89,108,116,190,194],"fundamental":[16],"challenge":[17],"for":[18,137,175],"modern":[19],"deep":[20],"generative":[21,191],"models.":[22],"Standard":[23],"Variational":[24],"Autoencoders":[25],"(VAEs)":[26],"employ":[27],"Gaussian":[28,46,113,173],"decoder":[29,55,114,188],"likelihoods":[30],"Lipschitz-constrained":[32],"neural":[33],"networks,":[34],"combination":[36],"that":[37,150,181],"is":[38],"structurally":[39],"incapable":[40],"of":[41,76,91,141,189],"producing":[42],"heavy-tailed":[43,146,176],"outputs:":[44],"the":[45,54,61,112,126,151,172,187,201],"tail":[47,92,155],"decays":[48],"exponentially,":[49],"Lipschitz":[51],"continuity":[52],"prevents":[53],"from":[56,60],"amplifying":[57],"rare":[58],"events":[59],"latent":[62,128],"space":[63],"input":[64],"to":[65,160,168,171,200],"sufficiently":[66],"overcome":[67],"this":[68,77],"decay.":[69],"We":[70],"provide":[71],"both":[72],"theoretical":[74],"characterization":[75],"limitation":[78],"controlled":[81],"empirical":[82],"demonstration":[83],"using":[84],"synthetic":[85],"Pareto":[86],"data":[87],"across":[88],"grid":[90],"indices":[93],"$\u03b1$":[94],"$\\in$":[95,103],"{2,":[96],"3,":[97],"5,":[98,105],"30}":[99],"dimensions":[101],"d":[102],"{1,":[104],"10}.":[106],"As":[107],"solution,":[109],"we":[110],"replace":[111],"with":[115],"Phase-Type":[117],"(PH)":[118],"distribution":[119],"based":[120],"on":[121],"Markov":[122,183],"chains,":[123],"while":[124],"keeping":[125],"encoder,":[127],"space,":[129],"training":[131],"procedure":[132],"identical.":[133],"PH":[134],"allow":[136],"arbitrarily":[138],"precise":[139],"approximations":[140],"any":[142],"positive-valued":[143],"distributions,":[144],"including":[145],"families.":[147],"Experiments":[148],"showed":[149],"PH-based":[152],"model":[153,192],"reduces":[154],"Kolmogorov-Smirnov":[156],"distance":[157],"by":[158,166],"up":[159,167],"x6":[161],"extreme":[163],"quantile":[164],"error":[165],"x10":[169],"compared":[170],"baseline":[174],"data.":[177],"These":[178],"results":[179],"demonstrate":[180],"integrating":[182],"chain-based":[184],"into":[186],"institutes":[193],"principled":[195],"practically":[197],"effective":[198],"solution":[199],"heavy-tail":[202],"generation":[203],"problem.":[204]},"counts_by_year":[],"updated_date":"2026-07-29T14:22:42.915294","created_date":"2026-05-21T00:00:00"}