{"id":"https://openalex.org/W7127514345","doi":"https://doi.org/10.48550/arxiv.2602.01591","title":"Know Your Step: Faster and Better Alignment for Flow Matching Models via Step-aware Advantages","display_name":"Know Your Step: Faster and Better Alignment for Flow Matching Models via Step-aware Advantages","publication_year":2026,"publication_date":"2026-02-02","ids":{"openalex":"https://openalex.org/W7127514345","doi":"https://doi.org/10.48550/arxiv.2602.01591"},"language":null,"primary_location":{"id":"pmh:doi:10.48550/arxiv.2602.01591","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":"publisher-specific-oa","license_id":"https://openalex.org/licenses/publisher-specific-oa","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/A5088311083","display_name":"Zhixiong Yue","orcid":"https://orcid.org/0000-0001-6269-5950"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yue, Zhixiong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5069102972","display_name":"Zixuan Ni","orcid":"https://orcid.org/0000-0001-5196-983X"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ni, Zixuan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5124940394","display_name":"Feiyang Ye","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ye, Feiyang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5124902235","display_name":"Jinshan Zhang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhang, Jinshan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5124887219","display_name":"Sheng Shen","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Shen, Sheng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5124928527","display_name":"Zhenpeng Mi","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Mi, Zhenpeng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":6,"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":0.5392000079154968,"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":0.5392000079154968,"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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.2295999974012375,"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/T11574","display_name":"Artificial Intelligence in Games","score":0.0957999974489212,"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/matching","display_name":"Matching (statistics)","score":0.5382000207901001},{"id":"https://openalex.org/keywords/noise","display_name":"Noise (video)","score":0.47110000252723694},{"id":"https://openalex.org/keywords/code","display_name":"Code (set theory)","score":0.43070000410079956},{"id":"https://openalex.org/keywords/reinforcement-learning","display_name":"Reinforcement learning","score":0.41990000009536743},{"id":"https://openalex.org/keywords/sampling","display_name":"Sampling (signal processing)","score":0.4189000129699707},{"id":"https://openalex.org/keywords/process","display_name":"Process (computing)","score":0.4106000065803528},{"id":"https://openalex.org/keywords/image","display_name":"Image (mathematics)","score":0.391400009393692},{"id":"https://openalex.org/keywords/noise-reduction","display_name":"Noise reduction","score":0.36320000886917114}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7477999925613403},{"id":"https://openalex.org/C165064840","wikidata":"https://www.wikidata.org/wiki/Q1321061","display_name":"Matching (statistics)","level":2,"score":0.5382000207901001},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5038999915122986},{"id":"https://openalex.org/C99498987","wikidata":"https://www.wikidata.org/wiki/Q2210247","display_name":"Noise (video)","level":3,"score":0.47110000252723694},{"id":"https://openalex.org/C2776760102","wikidata":"https://www.wikidata.org/wiki/Q5139990","display_name":"Code (set theory)","level":3,"score":0.43070000410079956},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.41990000009536743},{"id":"https://openalex.org/C140779682","wikidata":"https://www.wikidata.org/wiki/Q210868","display_name":"Sampling (signal processing)","level":3,"score":0.4189000129699707},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.4106000065803528},{"id":"https://openalex.org/C115961682","wikidata":"https://www.wikidata.org/wiki/Q860623","display_name":"Image (mathematics)","level":2,"score":0.391400009393692},{"id":"https://openalex.org/C163294075","wikidata":"https://www.wikidata.org/wiki/Q581861","display_name":"Noise reduction","level":2,"score":0.36320000886917114},{"id":"https://openalex.org/C126980161","wikidata":"https://www.wikidata.org/wiki/Q863783","display_name":"Simulated annealing","level":2,"score":0.3547999858856201},{"id":"https://openalex.org/C2781395549","wikidata":"https://www.wikidata.org/wiki/Q4680762","display_name":"Adaptive sampling","level":3,"score":0.3400000035762787},{"id":"https://openalex.org/C202615002","wikidata":"https://www.wikidata.org/wiki/Q783507","display_name":"Differentiable function","level":2,"score":0.33820000290870667},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.3352999985218048},{"id":"https://openalex.org/C38349280","wikidata":"https://www.wikidata.org/wiki/Q1434290","display_name":"Flow (mathematics)","level":2,"score":0.3215000033378601},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.31349998712539673},{"id":"https://openalex.org/C155542232","wikidata":"https://www.wikidata.org/wiki/Q736111","display_name":"Optical flow","level":3,"score":0.2969000041484833},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.29649999737739563},{"id":"https://openalex.org/C43126263","wikidata":"https://www.wikidata.org/wiki/Q128751","display_name":"Source code","level":2,"score":0.28439998626708984},{"id":"https://openalex.org/C73301696","wikidata":"https://www.wikidata.org/wiki/Q5469984","display_name":"Formalism (music)","level":3,"score":0.27900001406669617},{"id":"https://openalex.org/C22367795","wikidata":"https://www.wikidata.org/wiki/Q7625208","display_name":"Structured prediction","level":2,"score":0.27219998836517334},{"id":"https://openalex.org/C52740198","wikidata":"https://www.wikidata.org/wiki/Q1539564","display_name":"Importance sampling","level":3,"score":0.2612999975681305},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.25220000743865967}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:doi:10.48550/arxiv.2602.01591","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":"publisher-specific-oa","license_id":"https://openalex.org/licenses/publisher-specific-oa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Article"},{"id":"doi:10.48550/arxiv.2602.01591","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2602.01591","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":"article"}],"best_oa_location":{"id":"pmh:doi:10.48550/arxiv.2602.01591","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":"publisher-specific-oa","license_id":"https://openalex.org/licenses/publisher-specific-oa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Article"},"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":{"Recent":[0],"advances":[1],"in":[2,17,169],"flow":[3,30,76],"matching":[4,31,77],"models,":[5],"particularly":[6],"with":[7,64,89,184],"reinforcement":[8],"learning":[9],"(RL),":[10],"have":[11],"significantly":[12,177],"enhanced":[13],"human":[14,90,185],"preference":[15],"alignment":[16,180],"few":[18,84,170],"step":[19,62,85,104,131,153,171],"text":[20,78,172],"to":[21,50,79,139,173,197],"image":[22,80,174],"generators.":[23],"However,":[24],"existing":[25],"RL":[26],"based":[27],"approaches":[28],"for":[29,74,143,156],"models":[32,81,190],"typically":[33],"rely":[34],"on":[35],"numerous":[36],"denoising":[37],"steps,":[38],"while":[39,120],"suffering":[40],"from":[41],"sparse":[42],"and":[43,149,152,176,189],"imprecise":[44],"reward":[45,147],"signals":[46],"that":[47,163],"often":[48],"lead":[49],"suboptimal":[51],"alignment.":[52],"To":[53],"address":[54],"these":[55],"limitations,":[56],"we":[57],"propose":[58],"Temperature":[59],"Annealed":[60],"Few":[61],"Sampling":[63],"Group":[65],"Relative":[66],"Policy":[67],"Optimization":[68],"(TAFS":[69],"GRPO),":[70],"a":[71],"novel":[72],"framework":[73],"training":[75],"into":[82,116],"efficient":[83],"generators":[86],"well":[87],"aligned":[88],"preferences.":[91,186],"Our":[92],"method":[93],"iteratively":[94],"injects":[95],"adaptive":[96],"temporal":[97],"noise":[98],"onto":[99],"the":[100,109,117,122,137,141,144,179],"results":[101],"of":[102,125,146,181,191],"one":[103],"samples.":[105],"By":[106],"repeatedly":[107],"annealing":[108],"model's":[110],"sampled":[111],"outputs,":[112],"it":[113],"introduces":[114],"stochasticity":[115],"sampling":[118],"process":[119],"preserving":[121],"semantic":[123],"integrity":[124],"each":[126],"generated":[127,182],"image.":[128],"Moreover,":[129],"its":[130],"aware":[132],"advantage":[133],"integration":[134],"mechanism":[135],"combines":[136],"GRPO":[138,165],"avoid":[140],"need":[142],"differentiable":[145],"function":[148],"provide":[150],"dense":[151],"specific":[154],"rewards":[155],"stable":[157],"policy":[158],"optimization.":[159],"Extensive":[160],"experiments":[161],"demonstrate":[162],"TAFS":[164],"achieves":[166],"strong":[167],"performance":[168],"generation":[175],"improves":[178],"images":[183],"The":[187],"code":[188],"this":[192],"work":[193],"will":[194],"be":[195],"available":[196],"facilitate":[198],"further":[199],"research.":[200]},"counts_by_year":[],"updated_date":"2026-06-11T09:08:48.828518","created_date":"2026-02-04T00:00:00"}