{"id":"https://openalex.org/W7165398558","doi":"https://doi.org/10.48550/arxiv.2606.20404","title":"FlowBender: Feedback-Aware Training for Self-Correcting Conditional Flows","display_name":"FlowBender: Feedback-Aware Training for Self-Correcting Conditional Flows","publication_year":2026,"publication_date":"2026-06-18","ids":{"openalex":"https://openalex.org/W7165398558","doi":"https://doi.org/10.48550/arxiv.2606.20404"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2606.20404","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.20404","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.2606.20404","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5087876681","display_name":"Daniel Gilo","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Gilo, Daniel","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5005935318","display_name":"Sven Elflein","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Elflein, Sven","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139024795","display_name":"Ido Sobol","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Sobol, Ido","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5138999279","display_name":"Or Litany","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Litany, Or","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/T10775","display_name":"Generative Adversarial Networks and Image Synthesis","score":0.6636999845504761,"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":0.6636999845504761,"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/T10481","display_name":"Computer Graphics and Visualization Techniques","score":0.0786999985575676,"subfield":{"id":"https://openalex.org/subfields/1704","display_name":"Computer Graphics and Computer-Aided Design"},"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/T10719","display_name":"3D Shape Modeling and Analysis","score":0.07530000060796738,"subfield":{"id":"https://openalex.org/subfields/2206","display_name":"Computational Mechanics"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/fidelity","display_name":"Fidelity","score":0.6682000160217285},{"id":"https://openalex.org/keywords/differentiable-function","display_name":"Differentiable function","score":0.42480000853538513},{"id":"https://openalex.org/keywords/training","display_name":"Training (meteorology)","score":0.4196000099182129},{"id":"https://openalex.org/keywords/signal","display_name":"SIGNAL (programming language)","score":0.4009999930858612},{"id":"https://openalex.org/keywords/metric","display_name":"Metric (unit)","score":0.3779999911785126},{"id":"https://openalex.org/keywords/flow","display_name":"Flow (mathematics)","score":0.35190001130104065},{"id":"https://openalex.org/keywords/conditional-expectation","display_name":"Conditional expectation","score":0.336899995803833}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6995999813079834},{"id":"https://openalex.org/C2776459999","wikidata":"https://www.wikidata.org/wiki/Q2119376","display_name":"Fidelity","level":2,"score":0.6682000160217285},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4666000008583069},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.43860000371932983},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.42820000648498535},{"id":"https://openalex.org/C202615002","wikidata":"https://www.wikidata.org/wiki/Q783507","display_name":"Differentiable function","level":2,"score":0.42480000853538513},{"id":"https://openalex.org/C2777211547","wikidata":"https://www.wikidata.org/wiki/Q17141490","display_name":"Training (meteorology)","level":2,"score":0.4196000099182129},{"id":"https://openalex.org/C2779843651","wikidata":"https://www.wikidata.org/wiki/Q7390335","display_name":"SIGNAL (programming language)","level":2,"score":0.4009999930858612},{"id":"https://openalex.org/C176217482","wikidata":"https://www.wikidata.org/wiki/Q860554","display_name":"Metric (unit)","level":2,"score":0.3779999911785126},{"id":"https://openalex.org/C38349280","wikidata":"https://www.wikidata.org/wiki/Q1434290","display_name":"Flow (mathematics)","level":2,"score":0.35190001130104065},{"id":"https://openalex.org/C186215838","wikidata":"https://www.wikidata.org/wiki/Q772232","display_name":"Conditional expectation","level":2,"score":0.336899995803833},{"id":"https://openalex.org/C22367795","wikidata":"https://www.wikidata.org/wiki/Q7625208","display_name":"Structured prediction","level":2,"score":0.325300008058548},{"id":"https://openalex.org/C2778112365","wikidata":"https://www.wikidata.org/wiki/Q3511065","display_name":"Sequence (biology)","level":2,"score":0.30239999294281006},{"id":"https://openalex.org/C104267543","wikidata":"https://www.wikidata.org/wiki/Q208163","display_name":"Signal processing","level":3,"score":0.28049999475479126},{"id":"https://openalex.org/C103088060","wikidata":"https://www.wikidata.org/wiki/Q1062839","display_name":"Error detection and correction","level":2,"score":0.2694999873638153},{"id":"https://openalex.org/C44492722","wikidata":"https://www.wikidata.org/wiki/Q327069","display_name":"Conditional probability","level":2,"score":0.2606000006198883},{"id":"https://openalex.org/C155846161","wikidata":"https://www.wikidata.org/wiki/Q1143367","display_name":"Graphical model","level":2,"score":0.25099998712539673},{"id":"https://openalex.org/C113364801","wikidata":"https://www.wikidata.org/wiki/Q26674","display_name":"High fidelity","level":2,"score":0.250900000333786}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2606.20404","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.20404","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.2606.20404","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.20404","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":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Conditional":[0],"diffusion":[1],"and":[2,45,65,71,162,187,220,232,238],"flow":[3],"models":[4,55],"routinely":[5],"fail":[6],"to":[7,84,110,133,169],"satisfy":[8],"the":[9,29,33,39,58,85,88,91,97,105,131,150,159],"very":[10],"constraints":[11],"that":[12,56,74,96,104,122,206],"define":[13],"their":[14],"task.":[15],"For":[16,198],"instance,":[17],"a":[18,62,119,127,135,153,163,171,181,188,203,211],"depth-conditioned":[19],"model":[20,106],"often":[21],"produces":[22],"images":[23],"whose":[24],"re-extracted":[25],"depth":[26,36],"disagrees":[27],"with":[28],"input,":[30,129],"even":[31],"though":[32],"forward":[34,160],"operator--the":[35],"predictor":[37],"defining":[38],"constraint--is":[40],"available":[41],"during":[42],"both":[43,101],"training":[44,130],"inference.":[46],"Existing":[47],"approaches":[48],"generally":[49],"fall":[50],"into":[51],"two":[52],"categories:":[53],"supervised":[54,228],"treat":[57],"conditioning":[59],"signal":[60,168],"as":[61,126,195],"static":[63],"cue":[64],"ignore":[66],"alignment":[67,114],"information":[68],"at":[69,210],"inference,":[70],"guidance-based":[72],"methods":[73],"consult":[75],"it":[76],"through":[77],"hand-tuned":[78],"linear":[79],"updates,":[80],"typically":[81],"trading":[82,243],"fidelity":[83,237],"condition":[86],"against":[87,245],"plausibility":[89,239],"of":[90,178],"generated":[92],"sample.":[93],"We":[94,116,174],"argue":[95],"fundamental":[98],"gap":[99],"in":[100],"paradigms":[102],"is":[103,107,156],"never":[108],"trained":[109],"utilize":[111],"its":[112],"own":[113],"error.":[115],"introduce":[117,202],"FlowBender,":[118,179],"closed-loop":[120,208],"framework":[121],"treats":[123],"this":[124,167],"error":[125],"first-class":[128],"network":[132],"learn":[134],"correction":[136,209],"policy":[137],"conditioned":[138],"on":[139],"inference-time":[140,234],"feedback.":[141],"At":[142],"each":[143,246],"step,":[144],"an":[145],"unguided":[146],"look-ahead":[147],"pass":[148,165],"estimates":[149],"clean":[151],"signal,":[152],"task-specific":[154],"deviation":[155],"computed":[157],"via":[158],"operator,":[161],"refinement":[164],"consumes":[166],"produce":[170],"corrected":[172],"velocity.":[173],"propose":[175],"several":[176],"variants":[177],"including":[180],"gradient-based":[182],"formulation":[183],"for":[184,191],"differentiable":[185],"operators":[186],"zero-order":[189],"variant":[190],"non-differentiable":[192],"settings":[193],"such":[194],"JPEG":[196],"compression.":[197],"efficient":[199],"sampling,":[200],"we":[201],"prior-step":[204],"shortcut":[205],"enables":[207],"minimal":[212],"additional":[213],"computational":[214],"cost.":[215],"Across":[216],"image-to-image":[217],"translation,":[218],"restoration,":[219],"3D":[221],"mesh":[222],"texturing,":[223],"FlowBender":[224],"consistently":[225],"outperforms":[226],"standard":[227],"baselines,":[229],"alignment-loss-augmented":[230],"training,":[231],"state-of-the-art":[233],"guidance,":[235],"improving":[236],"simultaneously":[240],"rather":[241],"than":[242],"them":[244],"other.":[247],"Project":[248],"page:":[249],"https://flow-bender.github.io/":[250]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-06-20T00:00:00"}