{"id":"https://openalex.org/W7202390759","doi":"https://doi.org/10.48550/arxiv.2608.11544","title":"Fine-Tuning Generative Models for Extreme Events via CVaR-Penalized Wasserstein Gradient Flows","display_name":"Fine-Tuning Generative Models for Extreme Events via CVaR-Penalized Wasserstein Gradient Flows","publication_year":2026,"publication_date":"2026-08-12","ids":{"openalex":"https://openalex.org/W7202390759","doi":"https://doi.org/10.48550/arxiv.2608.11544"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2608.11544","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.11544","pdf_url":"https://arxiv.org/pdf/2608.11544","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,"listed_in":[],"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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"volume":null,"type":"preprint","is_traditional":true,"indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2608.11544","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5092912247","display_name":"Thejani Gamage","orcid":"https://orcid.org/0000-0002-9599-0873","observed_orcids":["https://orcid.org/0000-0002-9599-0873"]},"institutions":[{"id":"https://openalex.org/I24603500","display_name":"University of Massachusetts Amherst","ror":"https://ror.org/0072zz521","country_code":"US","type":"education","lineage":["https://openalex.org/I24603500"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Gamage, Thejani","raw_affiliation_strings":["Department of Mathematics and Statistics , University of Massachusetts Amherst , Amherst , MA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Mathematics and Statistics , University of Massachusetts Amherst , Amherst , MA","institution_ids":["https://openalex.org/I24603500"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5136148480","display_name":"Hyemin Gu","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I24603500","display_name":"University of Massachusetts Amherst","ror":"https://ror.org/0072zz521","country_code":"US","type":"education","lineage":["https://openalex.org/I24603500"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Gu, Hyemin","raw_affiliation_strings":["Department of Mathematics and Statistics , University of Massachusetts Amherst , Amherst , MA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Mathematics and Statistics , University of Massachusetts Amherst , Amherst , MA","institution_ids":["https://openalex.org/I24603500"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5147136294","display_name":"Zhizhen Zhang","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I24603500","display_name":"University of Massachusetts Amherst","ror":"https://ror.org/0072zz521","country_code":"US","type":"education","lineage":["https://openalex.org/I24603500"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Zhang, Zhizhen","raw_affiliation_strings":["Department of Mathematics and Statistics , University of Massachusetts Amherst , Amherst , MA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Mathematics and Statistics , University of Massachusetts Amherst , Amherst , MA","institution_ids":["https://openalex.org/I24603500"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5147081315","display_name":"Ziyu Chen","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chen, Ziyu","raw_affiliation_strings":["School of Data and Information Sciences (SDIS) and"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"School of Data and Information Sciences (SDIS) and","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5045079747","display_name":"Markos A. Katsoulakis","orcid":"https://orcid.org/0000-0003-4354-1766","observed_orcids":["https://orcid.org/0000-0003-4354-1766"]},"institutions":[{"id":"https://openalex.org/I24603500","display_name":"University of Massachusetts Amherst","ror":"https://ror.org/0072zz521","country_code":"US","type":"education","lineage":["https://openalex.org/I24603500"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Katsoulakis, Markos A.","raw_affiliation_strings":["Department of Mathematics and Statistics , University of Massachusetts Amherst , Amherst , MA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Mathematics and Statistics , University of Massachusetts Amherst , Amherst , MA","institution_ids":["https://openalex.org/I24603500"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5066349615","display_name":"Luc Rey-Bellet","orcid":"https://orcid.org/0000-0003-1166-8957","observed_orcids":["https://orcid.org/0000-0003-1166-8957"]},"institutions":[{"id":"https://openalex.org/I114027177","display_name":"University of North Carolina at Chapel Hill","ror":"https://ror.org/0130frc33","country_code":"US","type":"education","lineage":["https://openalex.org/I114027177"]},{"id":"https://openalex.org/I24603500","display_name":"University of Massachusetts Amherst","ror":"https://ror.org/0072zz521","country_code":"US","type":"education","lineage":["https://openalex.org/I24603500"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Rey-Bellet, Luc","raw_affiliation_strings":["Department of Mathematics , University of North Carolina at Chapel Hill , Chapel Hill , NC","Department of Mathematics and Statistics , University of Massachusetts Amherst , Amherst , MA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Mathematics , University of North Carolina at Chapel Hill , Chapel Hill , NC","institution_ids":["https://openalex.org/I114027177"]},{"raw_affiliation_string":"Department of Mathematics and Statistics , University of Massachusetts Amherst , Amherst , MA","institution_ids":["https://openalex.org/I24603500"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":2,"corresponding_author_ids":[],"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":0.11569999903440475,"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":0.11569999903440475,"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/T11206","display_name":"Model Reduction and Neural Networks","score":0.10939999669790268,"subfield":{"id":"https://openalex.org/subfields/3109","display_name":"Statistical and Nonlinear Physics"},"field":{"id":"https://openalex.org/fields/31","display_name":"Physics and Astronomy"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T11720","display_name":"Probability and Risk Models","score":0.10029999911785126,"subfield":{"id":"https://openalex.org/subfields/1803","display_name":"Management Science and Operations Research"},"field":{"id":"https://openalex.org/fields/18","display_name":"Decision Sciences"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/extreme-events","display_name":"extreme events","score":0.8240000009536743},{"id":"https://openalex.org/keywords/heavy-tailed-distributions","display_name":"heavy-tailed distributions","score":0.9610000252723694},{"id":"https://openalex.org/keywords/conditional-value-at-risk","display_name":"conditional value at risk","score":0.7689999938011169},{"id":"https://openalex.org/keywords/wasserstein-gradient-flow","display_name":"Wasserstein gradient flow","score":0.9620000123977661},{"id":"https://openalex.org/keywords/tail-risk","display_name":"tail risk","score":0.4440000057220459},{"id":"https://openalex.org/keywords/pre-trained-generative-models","display_name":"pre-trained generative models","score":0.6629999876022339},{"id":"https://openalex.org/keywords/diffusion-models","display_name":"diffusion models","score":0.6679999828338623},{"id":"https://openalex.org/keywords/gan","display_name":"GaN","score":0.8799999952316284},{"id":"https://openalex.org/keywords/streamflow","display_name":"river discharge","score":0.7239999771118164},{"id":"https://openalex.org/keywords/fama-french-portfolios","display_name":"Fama-French portfolios","score":0.9049999713897705},{"id":"https://openalex.org/keywords/tail-index","display_name":"tail index","score":0.3720000088214874}],"concepts":[{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.6581000089645386},{"id":"https://openalex.org/C22324862","wikidata":"https://www.wikidata.org/wiki/Q652707","display_name":"Lipschitz continuity","level":2,"score":0.6241999864578247},{"id":"https://openalex.org/C2779922397","wikidata":"https://www.wikidata.org/wiki/Q5014755","display_name":"CVAR","level":4,"score":0.5918999910354614},{"id":"https://openalex.org/C207390915","wikidata":"https://www.wikidata.org/wiki/Q1230525","display_name":"Divergence (linguistics)","level":2,"score":0.5295000076293945},{"id":"https://openalex.org/C34388435","wikidata":"https://www.wikidata.org/wiki/Q2267362","display_name":"Bounded function","level":2,"score":0.5187000036239624},{"id":"https://openalex.org/C38349280","wikidata":"https://www.wikidata.org/wiki/Q1434290","display_name":"Flow (mathematics)","level":2,"score":0.49889999628067017},{"id":"https://openalex.org/C126255220","wikidata":"https://www.wikidata.org/wiki/Q141495","display_name":"Mathematical optimization","level":1,"score":0.48660001158714294},{"id":"https://openalex.org/C28826006","wikidata":"https://www.wikidata.org/wiki/Q33521","display_name":"Applied mathematics","level":1,"score":0.4388999938964844},{"id":"https://openalex.org/C185429906","wikidata":"https://www.wikidata.org/wiki/Q1130160","display_name":"Estimator","level":2,"score":0.41280001401901245},{"id":"https://openalex.org/C184050105","wikidata":"https://www.wikidata.org/wiki/Q273163","display_name":"Isotropy","level":2,"score":0.38670000433921814},{"id":"https://openalex.org/C107673813","wikidata":"https://www.wikidata.org/wiki/Q812534","display_name":"Bayesian probability","level":2,"score":0.38280001282691956},{"id":"https://openalex.org/C177264268","wikidata":"https://www.wikidata.org/wiki/Q1514741","display_name":"Set (abstract data type)","level":2,"score":0.3752000033855438},{"id":"https://openalex.org/C167879884","wikidata":"https://www.wikidata.org/wiki/Q727568","display_name":"Balanced flow","level":2,"score":0.3352000117301941},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.326200008392334},{"id":"https://openalex.org/C167966045","wikidata":"https://www.wikidata.org/wiki/Q5532625","display_name":"Generative model","level":3,"score":0.3237999975681305},{"id":"https://openalex.org/C2775936607","wikidata":"https://www.wikidata.org/wiki/Q466845","display_name":"Tracking (education)","level":2,"score":0.31459999084472656},{"id":"https://openalex.org/C151201525","wikidata":"https://www.wikidata.org/wiki/Q177239","display_name":"Limit (mathematics)","level":2,"score":0.3140999972820282},{"id":"https://openalex.org/C153258448","wikidata":"https://www.wikidata.org/wiki/Q1199743","display_name":"Gradient descent","level":3,"score":0.30480000376701355},{"id":"https://openalex.org/C57691317","wikidata":"https://www.wikidata.org/wiki/Q1289248","display_name":"Scalar (mathematics)","level":2,"score":0.2980000078678131},{"id":"https://openalex.org/C159886148","wikidata":"https://www.wikidata.org/wiki/Q176645","display_name":"Markov process","level":2,"score":0.29190000891685486},{"id":"https://openalex.org/C112680207","wikidata":"https://www.wikidata.org/wiki/Q714886","display_name":"Regular polygon","level":2,"score":0.2858999967575073},{"id":"https://openalex.org/C207201462","wikidata":"https://www.wikidata.org/wiki/Q182505","display_name":"Bayes' theorem","level":3,"score":0.28040000796318054},{"id":"https://openalex.org/C49937458","wikidata":"https://www.wikidata.org/wiki/Q2599292","display_name":"Probabilistic logic","level":2,"score":0.2648000121116638},{"id":"https://openalex.org/C2780009758","wikidata":"https://www.wikidata.org/wiki/Q6804172","display_name":"Measure (data warehouse)","level":2,"score":0.2628999948501587},{"id":"https://openalex.org/C134306372","wikidata":"https://www.wikidata.org/wiki/Q7754","display_name":"Mathematical analysis","level":1,"score":0.25920000672340393},{"id":"https://openalex.org/C159176650","wikidata":"https://www.wikidata.org/wiki/Q43261","display_name":"Horizon","level":2,"score":0.258899986743927},{"id":"https://openalex.org/C21031990","wikidata":"https://www.wikidata.org/wiki/Q355020","display_name":"Probability measure","level":2,"score":0.2556999921798706},{"id":"https://openalex.org/C2778112365","wikidata":"https://www.wikidata.org/wiki/Q3511065","display_name":"Sequence (biology)","level":2,"score":0.2540999948978424},{"id":"https://openalex.org/C52740198","wikidata":"https://www.wikidata.org/wiki/Q1539564","display_name":"Importance sampling","level":3,"score":0.2531999945640564},{"id":"https://openalex.org/C99498987","wikidata":"https://www.wikidata.org/wiki/Q2210247","display_name":"Noise (video)","level":3,"score":0.2531999945640564},{"id":"https://openalex.org/C145071142","wikidata":"https://www.wikidata.org/wiki/Q1411116","display_name":"Fictitious play","level":3,"score":0.25220000743865967}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2608.11544","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.11544","pdf_url":"https://arxiv.org/pdf/2608.11544","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,"listed_in":[],"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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2608.11544","doi":"https://doi.org/10.48550/arxiv.2608.11544","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2608.11544","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,"listed_in":[],"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"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2608.11544","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.11544","pdf_url":"https://arxiv.org/pdf/2608.11544","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,"listed_in":[],"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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"study_designs":[],"awards":[{"id":"https://openalex.org/G1523888516","display_name":null,"funder_award_id":"FA9550-","funder_id":"https://openalex.org/F4320338279","funder_display_name":"Air Force Office of Scientific Research"},{"id":"https://openalex.org/G4516736450","display_name":null,"funder_award_id":"-21-1-","funder_id":"https://openalex.org/F4320338279","funder_display_name":"Air Force Office of Scientific Research"},{"id":"https://openalex.org/G6478756169","display_name":null,"funder_award_id":"FA9550-21","funder_id":"https://openalex.org/F4320338279","funder_display_name":"Air Force Office of Scientific Research"},{"id":"https://openalex.org/G6518258739","display_name":null,"funder_award_id":"FA9550-21-1-0354","funder_id":"https://openalex.org/F4320338279","funder_display_name":"Air Force Office of Scientific Research"},{"id":"https://openalex.org/G8127691482","display_name":"Regularized divergences and their gradient flows, generative modeling and structure-preserving learning.","funder_award_id":"2307115","funder_id":"https://openalex.org/F4320306076","funder_display_name":"National Science Foundation"}],"funders":[{"id":"https://openalex.org/F4320306076","display_name":"National Science Foundation","ror":"https://ror.org/021nxhr62"},{"id":"https://openalex.org/F4320338279","display_name":"Air Force Office of Scientific Research","ror":"https://ror.org/011e9bt93"}],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7202390759.pdf","grobid_xml":"https://content.openalex.org/works/W7202390759.grobid-xml"},"referenced_works_count":6,"referenced_works":["https://openalex.org/W1506131801","https://openalex.org/W2169314767","https://openalex.org/W2785678896","https://openalex.org/W4400064524","https://openalex.org/W4404336892","https://openalex.org/W7130732484"],"related_works":["https://openalex.org/W7130732484","https://openalex.org/W4404336892","https://openalex.org/W4394705186","https://openalex.org/W7221756343","https://openalex.org/W7161766341","https://openalex.org/W4405939974","https://openalex.org/W7201941608","https://openalex.org/W7204635051","https://openalex.org/W7164197626","https://openalex.org/W4283331950"],"abstract_inverted_index":{"In":[0],"many":[1],"high-stakes":[2],"domains,":[3],"extreme":[4,31],"events":[5],"carry":[6],"substantial":[7],"consequences,":[8],"yet":[9],"learning":[10],"the":[11,22,34,50,58,82,87,104,108,116,120,127,140,143,147,150,157,173,188,191,197,213,219],"heavy-tailed":[12,74],"distributions":[13],"that":[14,169],"govern":[15],"them":[16],"from":[17,103,227],"finite":[18],"samples":[19,106],"remains":[20,170],"challenging:":[21],"quantities":[23],"of":[24,81,86,107,119,190,242,265],"interest":[25],"are":[26,112],"driven":[27],"by":[28,92,239],"a":[29,66,78,93,98,161,166,178,262],"few":[30],"observations,":[32],"so":[33],"tail":[35,51,163,175,224,237],"is":[36],"under-sampled":[37,174],"relative":[38],"to":[39,126,196,229],"its":[40],"importance.":[41],"Even":[42],"generative":[43,71,259],"models":[44,72,252],"tailored":[45],"for":[46,69],"heavy":[47],"tails":[48],"capture":[49],"region":[52,176],"inadequately":[53],"in":[54,172,212],"practice.":[55],"We":[56],"propose":[57],"Conditional":[59],"Value-at-Risk":[60],"(CVaR)-penalized":[61],"Generative":[62],"Particle":[63],"Algorithm":[64],"(CVaR-GPA),":[65],"tail-agnostic":[67],"algorithm":[68],"fine-tuning":[70,231],"toward":[73],"targets,":[75,204],"built":[76],"as":[77],"time":[79],"discretization":[80],"Wasserstein":[83],"gradient":[84,117],"flow":[85,99,148],"Lipschitz-regularized":[88,133],"KL":[89,134],"divergence":[90,135],"penalized":[91],"CVaR":[94,144,153],"discrepancy":[95],"term.":[96],"Such":[97],"can":[100],"be":[101],"initialized":[102],"output":[105],"pre-trained":[109,128,251],"model,":[110],"which":[111],"then":[113],"transported":[114],"along":[115],"descent":[118],"loss":[121],"functional,":[122],"without":[123,199],"requiring":[124],"access":[125],"model's":[129],"internal":[130],"architecture.":[131],"The":[132,152,182],"requires":[136],"minimal":[137],"assumptions":[138],"on":[139,149,156],"target,":[141,198],"while":[142],"penalty":[145,154],"focuses":[146],"tail.":[151],"depends":[155],"target":[158],"only":[159],"through":[160],"scalar":[162],"statistic,":[164],"inducing":[165],"velocity":[167,184],"field":[168],"active":[171],"at":[177],"dimension-free":[179],"estimation":[180],"cost.":[181],"resulting":[183],"field,":[185],"and":[186,218,236,245,257],"hence":[187],"depth":[189],"transport":[192],"map,":[193],"implicitly":[194],"adapts":[195],"target-specific":[200],"modifications.":[201],"Across":[202],"four":[203],"including":[205],"two":[206],"real-world,":[207],"high-dimensional":[208],"datasets":[209],"(daily":[210],"streamflow":[211],"Ohio":[214],"River":[215],"basin":[216],"($d=64$)":[217],"Fama-French":[220],"portfolios":[221],"($d=25$)":[222],"with":[223,232,261],"indices":[225],"ranging":[226],"$1.05$":[228],"$3.34$),":[230],"CVaR-GPA":[233],"reduces":[234],"global":[235],"errors":[238],"geometric-mean":[240],"factors":[241],"$14.0":[243],"\\times$":[244],"$9.8":[246],"\\times$,":[247],"respectively,":[248],"across":[249],"seven":[250],"spanning":[253],"GANs,":[254],"diffusion":[255],"models,":[256],"other":[258],"flows,":[260],"single":[263],"set":[264],"hyperparameters.":[266]},"counts_by_year":[],"updated_date":"2026-11-08T12:12:28.789561","created_date":"2026-08-14T00:00:00"}