{"id":"https://openalex.org/W7133329671","doi":"https://doi.org/10.48550/arxiv.2603.01563","title":"LFPO: Likelihood-Free Policy Optimization for Masked Diffusion Models","display_name":"LFPO: Likelihood-Free Policy Optimization for Masked Diffusion Models","publication_year":2026,"publication_date":"2026-03-02","ids":{"openalex":"https://openalex.org/W7133329671","doi":"https://doi.org/10.48550/arxiv.2603.01563"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2603.01563","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.01563","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.2603.01563","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5008764343","display_name":"C.-H. Wei","orcid":"https://orcid.org/0009-0003-8400-0138"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wei, Chenxing","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5127950856","display_name":"Jiazhen Kang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Kang, Jiazhen","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5128012711","display_name":"Hong Wang (13528)","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Hong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5056834769","display_name":"Jianqing Zhang","orcid":"https://orcid.org/0009-0003-4990-8466"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhang, Jianqing","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5127923890","display_name":"Hao Jiang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jiang, Hao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5127930973","display_name":"Xiaolong Xu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xu, Xiaolong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5127920436","display_name":"Ningyuan Sun","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Sun, Ningyuan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5127904962","display_name":"Ying He","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"He, Ying","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5127974786","display_name":"F. Richard Yu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yu, F. Richard","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5127922512","display_name":"Yao Shu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Shu, Yao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5128031181","display_name":"Bo Jiang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jiang, Bo","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":0.16670000553131104,"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":0.16670000553131104,"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/T10462","display_name":"Reinforcement Learning in Robotics","score":0.11249999701976776,"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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.08720000088214874,"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/correctness","display_name":"Correctness","score":0.6456999778747559},{"id":"https://openalex.org/keywords/consistency","display_name":"Consistency (knowledge bases)","score":0.6039999723434448},{"id":"https://openalex.org/keywords/field","display_name":"Field (mathematics)","score":0.46309998631477356},{"id":"https://openalex.org/keywords/security-token","display_name":"Security token","score":0.4352000057697296},{"id":"https://openalex.org/keywords/code","display_name":"Code (set theory)","score":0.4318000078201294},{"id":"https://openalex.org/keywords/estimator","display_name":"Estimator","score":0.4262000024318695},{"id":"https://openalex.org/keywords/matching","display_name":"Matching (statistics)","score":0.41690000891685486},{"id":"https://openalex.org/keywords/structured-prediction","display_name":"Structured prediction","score":0.37369999289512634},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.3677000105381012}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6733999848365784},{"id":"https://openalex.org/C55439883","wikidata":"https://www.wikidata.org/wiki/Q360812","display_name":"Correctness","level":2,"score":0.6456999778747559},{"id":"https://openalex.org/C2776436953","wikidata":"https://www.wikidata.org/wiki/Q5163215","display_name":"Consistency (knowledge bases)","level":2,"score":0.6039999723434448},{"id":"https://openalex.org/C9652623","wikidata":"https://www.wikidata.org/wiki/Q190109","display_name":"Field (mathematics)","level":2,"score":0.46309998631477356},{"id":"https://openalex.org/C48145219","wikidata":"https://www.wikidata.org/wiki/Q1335365","display_name":"Security token","level":2,"score":0.4352000057697296},{"id":"https://openalex.org/C2776760102","wikidata":"https://www.wikidata.org/wiki/Q5139990","display_name":"Code (set theory)","level":3,"score":0.4318000078201294},{"id":"https://openalex.org/C185429906","wikidata":"https://www.wikidata.org/wiki/Q1130160","display_name":"Estimator","level":2,"score":0.4262000024318695},{"id":"https://openalex.org/C126255220","wikidata":"https://www.wikidata.org/wiki/Q141495","display_name":"Mathematical optimization","level":1,"score":0.4235999882221222},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.42010000348091125},{"id":"https://openalex.org/C165064840","wikidata":"https://www.wikidata.org/wiki/Q1321061","display_name":"Matching (statistics)","level":2,"score":0.41690000891685486},{"id":"https://openalex.org/C22367795","wikidata":"https://www.wikidata.org/wiki/Q7625208","display_name":"Structured prediction","level":2,"score":0.37369999289512634},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.3677000105381012},{"id":"https://openalex.org/C92757383","wikidata":"https://www.wikidata.org/wiki/Q382497","display_name":"Affine transformation","level":2,"score":0.367000013589859},{"id":"https://openalex.org/C38349280","wikidata":"https://www.wikidata.org/wiki/Q1434290","display_name":"Flow (mathematics)","level":2,"score":0.36500000953674316},{"id":"https://openalex.org/C205203396","wikidata":"https://www.wikidata.org/wiki/Q612143","display_name":"Bilinear interpolation","level":2,"score":0.34599998593330383},{"id":"https://openalex.org/C85847156","wikidata":"https://www.wikidata.org/wiki/Q59015987","display_name":"Verifiable secret sharing","level":3,"score":0.3449999988079071},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.34439998865127563},{"id":"https://openalex.org/C167879884","wikidata":"https://www.wikidata.org/wiki/Q727568","display_name":"Balanced flow","level":2,"score":0.33880001306533813},{"id":"https://openalex.org/C111335779","wikidata":"https://www.wikidata.org/wiki/Q3454686","display_name":"Reduction (mathematics)","level":2,"score":0.31869998574256897},{"id":"https://openalex.org/C159877910","wikidata":"https://www.wikidata.org/wiki/Q2202883","display_name":"Autoregressive model","level":2,"score":0.31139999628067017},{"id":"https://openalex.org/C147764199","wikidata":"https://www.wikidata.org/wiki/Q6865248","display_name":"Minification","level":2,"score":0.30219998955726624},{"id":"https://openalex.org/C43126263","wikidata":"https://www.wikidata.org/wiki/Q128751","display_name":"Source code","level":2,"score":0.29429998993873596},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.2930999994277954},{"id":"https://openalex.org/C137836250","wikidata":"https://www.wikidata.org/wiki/Q984063","display_name":"Optimization problem","level":2,"score":0.287200003862381},{"id":"https://openalex.org/C69357855","wikidata":"https://www.wikidata.org/wiki/Q163214","display_name":"Diffusion","level":2,"score":0.26600000262260437},{"id":"https://openalex.org/C121099081","wikidata":"https://www.wikidata.org/wiki/Q665580","display_name":"Trap (plumbing)","level":2,"score":0.25999999046325684}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2603.01563","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.01563","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.2603.01563","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.01563","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":{"Reinforcement":[0],"Learning":[1],"with":[2,133],"Verifiable":[3],"Rewards":[4],"(RLVR)":[5],"has":[6],"achieved":[7],"remarkable":[8],"success":[9],"in":[10,15,105],"improving":[11],"autoregressive":[12],"models,":[13],"especially":[14],"domains":[16],"requiring":[17],"correctness":[18],"like":[19],"mathematical":[20],"reasoning":[21,150],"and":[22,149],"code":[23,148],"generation.":[24],"However,":[25],"directly":[26,91],"applying":[27],"such":[28],"paradigms":[29],"to":[30,50,77,129],"Diffusion":[31],"Large":[32],"Language":[33],"Models":[34],"(dLLMs)":[35],"is":[36],"fundamentally":[37],"hindered":[38],"by":[39,117,156],"the":[40,70,78,102,109,126],"intractability":[41],"of":[42,72],"exact":[43],"likelihood":[44,106],"computation,":[45],"which":[46,90],"forces":[47],"existing":[48],"methods":[49],"rely":[51],"on":[52,147],"high-variance":[53],"approximations.":[54],"To":[55],"bridge":[56],"this":[57],"gap,":[58],"we":[59],"propose":[60],"Likelihood-Free":[61],"Policy":[62],"Optimization":[63],"(LFPO),":[64],"a":[65],"native":[66],"framework":[67],"that":[68,140],"maps":[69],"concept":[71],"vector":[73],"field":[74],"flow":[75,128],"matching":[76],"discrete":[79],"token":[80],"space.":[81],"Specifically,":[82],"LFPO":[83,114,141],"formulates":[84],"alignment":[85],"as":[86],"geometric":[87],"velocity":[88],"rectification,":[89],"optimizes":[92],"denoising":[93],"logits":[94],"via":[95],"contrastive":[96],"updates.":[97],"This":[98],"design":[99],"effectively":[100,124],"bypasses":[101],"errors":[103],"inherent":[104],"approximation,":[107],"yielding":[108],"precise":[110],"gradient":[111],"estimation.":[112],"Furthermore,":[113],"enforce":[115],"consistency":[116],"predicting":[118],"final":[119],"solutions":[120],"from":[121],"intermediate":[122],"steps,":[123],"straightening":[125],"probability":[127],"enable":[130],"high-quality":[131],"generation":[132],"significantly":[134],"fewer":[135],"iterations.":[136],"Extensive":[137],"experiments":[138],"demonstrate":[139],"not":[142],"only":[143],"outperforms":[144],"state-of-the-art":[145],"baselines":[146],"benchmarks":[151],"but":[152],"also":[153],"accelerates":[154],"inference":[155],"approximately":[157],"20%":[158],"through":[159],"reduced":[160],"diffusion":[161],"steps.":[162]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-03-04T00:00:00"}