{"id":"https://openalex.org/W7160897818","doi":"https://doi.org/10.48550/arxiv.2605.08515","title":"Quantile-Coupled Flow Matching for Distributional Reinforcement Learning","display_name":"Quantile-Coupled Flow Matching for Distributional Reinforcement Learning","publication_year":2026,"publication_date":"2026-05-08","ids":{"openalex":"https://openalex.org/W7160897818","doi":"https://doi.org/10.48550/arxiv.2605.08515"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2605.08515","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.08515","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":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.2605.08515","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5063059819","display_name":"Michael Groom","orcid":"https://orcid.org/0000-0003-2473-7229"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Groom, Michael","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5028344749","display_name":"Victor-Alexandru Darvariu","orcid":"https://orcid.org/0000-0001-9250-8175"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Darvariu, Victor-Alexandru","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135935626","display_name":"Lars Kunze","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Kunze, Lars","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5128242078","display_name":"James Wilson","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wilson, James","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5135995763","display_name":"Nick Hawes","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Hawes, Nick","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/T10462","display_name":"Reinforcement Learning in Robotics","score":0.5127000212669373,"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/T10462","display_name":"Reinforcement Learning in Robotics","score":0.5127000212669373,"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/T11206","display_name":"Model Reduction and Neural Networks","score":0.18050000071525574,"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/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.11819999665021896,"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/reinforcement-learning","display_name":"Reinforcement learning","score":0.7297999858856201},{"id":"https://openalex.org/keywords/matching","display_name":"Matching (statistics)","score":0.6539999842643738},{"id":"https://openalex.org/keywords/metric","display_name":"Metric (unit)","score":0.5748000144958496},{"id":"https://openalex.org/keywords/projection","display_name":"Projection (relational algebra)","score":0.565500020980835},{"id":"https://openalex.org/keywords/monotone-polygon","display_name":"Monotone polygon","score":0.49790000915527344},{"id":"https://openalex.org/keywords/flow","display_name":"Flow (mathematics)","score":0.39649999141693115},{"id":"https://openalex.org/keywords/dynamic-programming","display_name":"Dynamic programming","score":0.364300012588501},{"id":"https://openalex.org/keywords/bellman-equation","display_name":"Bellman equation","score":0.3529999852180481}],"concepts":[{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.7297999858856201},{"id":"https://openalex.org/C165064840","wikidata":"https://www.wikidata.org/wiki/Q1321061","display_name":"Matching (statistics)","level":2,"score":0.6539999842643738},{"id":"https://openalex.org/C176217482","wikidata":"https://www.wikidata.org/wiki/Q860554","display_name":"Metric (unit)","level":2,"score":0.5748000144958496},{"id":"https://openalex.org/C57493831","wikidata":"https://www.wikidata.org/wiki/Q3134666","display_name":"Projection (relational algebra)","level":2,"score":0.565500020980835},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5325000286102295},{"id":"https://openalex.org/C2834757","wikidata":"https://www.wikidata.org/wiki/Q4925424","display_name":"Monotone polygon","level":2,"score":0.49790000915527344},{"id":"https://openalex.org/C126255220","wikidata":"https://www.wikidata.org/wiki/Q141495","display_name":"Mathematical optimization","level":1,"score":0.483599990606308},{"id":"https://openalex.org/C38349280","wikidata":"https://www.wikidata.org/wiki/Q1434290","display_name":"Flow (mathematics)","level":2,"score":0.39649999141693115},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.3671000003814697},{"id":"https://openalex.org/C37404715","wikidata":"https://www.wikidata.org/wiki/Q380679","display_name":"Dynamic programming","level":2,"score":0.364300012588501},{"id":"https://openalex.org/C14646407","wikidata":"https://www.wikidata.org/wiki/Q1430750","display_name":"Bellman equation","level":2,"score":0.3529999852180481},{"id":"https://openalex.org/C2164484","wikidata":"https://www.wikidata.org/wiki/Q5170150","display_name":"Core (optical fiber)","level":2,"score":0.3481000065803528},{"id":"https://openalex.org/C17020691","wikidata":"https://www.wikidata.org/wiki/Q139677","display_name":"Operator (biology)","level":5,"score":0.3319000005722046},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.2921999990940094},{"id":"https://openalex.org/C72169020","wikidata":"https://www.wikidata.org/wiki/Q194404","display_name":"Monotonic function","level":2,"score":0.28949999809265137},{"id":"https://openalex.org/C2776502983","wikidata":"https://www.wikidata.org/wiki/Q690182","display_name":"Contrast (vision)","level":2,"score":0.26899999380111694},{"id":"https://openalex.org/C2780898871","wikidata":"https://www.wikidata.org/wiki/Q860554","display_name":"Performance metric","level":2,"score":0.2639000117778778},{"id":"https://openalex.org/C123853557","wikidata":"https://www.wikidata.org/wiki/Q7098946","display_name":"Optimal matching","level":3,"score":0.25769999623298645},{"id":"https://openalex.org/C2777634741","wikidata":"https://www.wikidata.org/wiki/Q768993","display_name":"Wasserstein metric","level":2,"score":0.2565999925136566},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.2524999976158142},{"id":"https://openalex.org/C186215838","wikidata":"https://www.wikidata.org/wiki/Q772232","display_name":"Conditional expectation","level":2,"score":0.2524000108242035}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2605.08515","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.08515","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":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.2605.08515","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.08515","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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"sustainable_development_goals":[{"display_name":"Peace, Justice and strong institutions","score":0.8147146701812744,"id":"https://metadata.un.org/sdg/16"}],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Unlike":[0],"standard":[1],"expected-return":[2],"Reinforcement":[3],"Learning":[4],"(RL),":[5],"Distributional":[6],"RL":[7,188],"(DRL)":[8],"models":[9,164],"the":[10,50,57,109,124,138,147],"full":[11],"return":[12,35,82],"distribution,":[13],"making":[14],"it":[15],"better-suited":[16],"for":[17,31,78,165],"uncertainty-aware":[18],"and":[19,100],"risk-sensitive":[20],"decision-making.":[21],"Conditional":[22],"Flow":[23],"Matching":[24],"(CFM)":[25],"critics":[26,63],"have":[27],"recently":[28],"attracted":[29],"attention":[30],"modelling":[32],"continuous,":[33],"multi-modal":[34],"distributions.":[36,83],"Despite":[37],"this":[38,85,89],"interest,":[39],"there":[40],"remains":[41],"a":[42,94,132,196],"substantial":[43],"metric":[44],"mismatch:":[45],"DRL":[46,206],"theory":[47],"relies":[48],"on":[49,186],"distributional":[51,150],"Bellman":[52,80,101],"operator":[53],"being":[54],"contractive":[55],"in":[56],"$p$-Wasserstein":[58],"distance,":[59],"yet":[60],"existing":[61],"CFM":[62,95,129,179,199],"are":[64,74],"trained":[65],"with":[66,117,137,152,162,205],"arbitrary":[67,115],"source-target":[68],"couplings,":[69],"so":[70],"their":[71],"flow-matching":[72,149],"losses":[73],"not":[75],"Wasserstein-aligned":[76,133,155],"surrogates":[77],"matching":[79],"target":[81,102],"In":[84],"work,":[86],"we":[87],"address":[88],"mismatch":[90],"by":[91],"proposing":[92],"FlowIQN,":[93],"critic":[96,130,151,200],"that":[97,123,171,201],"sorts":[98],"source":[99],"samples":[103],"within":[104],"each":[105],"mini-batch":[106],"to":[107],"approximate":[108,134],"monotone":[110],"optimal":[111],"transport":[112],"coupling,":[113],"replacing":[114],"pairings":[116],"quantile-aligned":[118],"flow":[119],"paths.":[120],"We":[121,158],"prove":[122],"loss":[125],"of":[126,140],"our":[127,143],"quantile-coupled":[128],"yields":[131,183],"projection":[135,156],"compatible":[136,204],"foundations":[139],"DRL.":[141],"To":[142],"knowledge,":[144],"FlowIQN":[145,161,172],"is":[146,202],"first":[148],"an":[153],"explicit":[154],"guarantee.":[157],"further":[159],"extend":[160],"shortcut":[163],"efficient":[166],"inference.":[167],"Empirical":[168],"results":[169],"show":[170],"improves":[173],"Wasserstein":[174],"return-distribution":[175],"accuracy":[176],"over":[177],"other":[178],"critics.":[180],"It":[181],"also":[182],"competitive":[184],"performance":[185],"offline":[187],"benchmarks":[189],"across":[190],"multiple":[191],"policy":[192],"extraction":[193],"methods,":[194],"providing":[195],"theoretically":[197],"grounded":[198],"readily":[203],"pipelines.":[207],"Code:":[208],"https://github.com/ori-goals/flowIQN.":[209]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-05-13T00:00:00"}