{"id":"https://openalex.org/W7214778286","doi":"https://doi.org/10.48550/arxiv.2609.32333","title":"Progressive-View On-Policy Distillation for Regional-to-Global Transfer in Multimodal LLMs","display_name":"Progressive-View On-Policy Distillation for Regional-to-Global Transfer in Multimodal LLMs","publication_year":2026,"publication_date":"2026-09-26","ids":{"openalex":"https://openalex.org/W7214778286","doi":"https://doi.org/10.48550/arxiv.2609.32333"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2609.32333","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.32333","pdf_url":"https://arxiv.org/pdf/2609.32333","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/2609.32333","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5153410258","display_name":"Shanfeng Huang","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I98301712","display_name":"Baidu (China)","ror":"https://ror.org/03vs3wt56","country_code":"CN","type":"company","lineage":["https://openalex.org/I98301712"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Huang, Shanfeng","raw_affiliation_strings":["Baidu Inc"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Baidu Inc","institution_ids":["https://openalex.org/I98301712"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5153542870","display_name":"Zhou Fang","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I98301712","display_name":"Baidu (China)","ror":"https://ror.org/03vs3wt56","country_code":"CN","type":"company","lineage":["https://openalex.org/I98301712"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Fang, Zhou","raw_affiliation_strings":["Baidu Inc"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Baidu Inc","institution_ids":["https://openalex.org/I98301712"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5153546894","display_name":"Song Xiao","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I98301712","display_name":"Baidu (China)","ror":"https://ror.org/03vs3wt56","country_code":"CN","type":"company","lineage":["https://openalex.org/I98301712"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Xiao, Song","raw_affiliation_strings":["Baidu Inc"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Baidu Inc","institution_ids":["https://openalex.org/I98301712"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5153403532","display_name":"Hai Du","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I98301712","display_name":"Baidu (China)","ror":"https://ror.org/03vs3wt56","country_code":"CN","type":"company","lineage":["https://openalex.org/I98301712"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Du, Hai","raw_affiliation_strings":["Baidu Inc"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Baidu Inc","institution_ids":["https://openalex.org/I98301712"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I98301712"],"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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.9984999895095825,"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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.9984999895095825,"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/T10036","display_name":"Advanced Neural Network Applications","score":0.0007999999797903001,"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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.00039999998989515007,"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/on-policy-distillation","display_name":"on-policy distillation","score":0.7549999952316284},{"id":"https://openalex.org/keywords/jensen-shannon-divergence","display_name":"Jensen-Shannon divergence","score":0.8240000009536743},{"id":"https://openalex.org/keywords/multimodal-large-language-models","display_name":"multimodal large language models","score":0.8579999804496765},{"id":"https://openalex.org/keywords/visual-mathematics","display_name":"visual mathematics","score":0.5559999942779541},{"id":"https://openalex.org/keywords/multimodal-question-answering","display_name":"multimodal question answering","score":0.8009999990463257}],"concepts":[{"id":"https://openalex.org/C204030448","wikidata":"https://www.wikidata.org/wiki/Q101017","display_name":"Distillation","level":2,"score":0.73089998960495},{"id":"https://openalex.org/C176217482","wikidata":"https://www.wikidata.org/wiki/Q860554","display_name":"Metric (unit)","level":2,"score":0.6948999762535095},{"id":"https://openalex.org/C183115368","wikidata":"https://www.wikidata.org/wiki/Q856577","display_name":"Weighting","level":2,"score":0.6930999755859375},{"id":"https://openalex.org/C165064840","wikidata":"https://www.wikidata.org/wiki/Q1321061","display_name":"Matching (statistics)","level":2,"score":0.5997999906539917},{"id":"https://openalex.org/C774472","wikidata":"https://www.wikidata.org/wiki/Q6760393","display_name":"Margin (machine learning)","level":2,"score":0.57669997215271},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5644999742507935},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4941999912261963},{"id":"https://openalex.org/C124681953","wikidata":"https://www.wikidata.org/wiki/Q339062","display_name":"Decomposition","level":2,"score":0.48649999499320984},{"id":"https://openalex.org/C68387754","wikidata":"https://www.wikidata.org/wiki/Q7271585","display_name":"Schedule","level":2,"score":0.4616999924182892},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.4253999888896942},{"id":"https://openalex.org/C12725497","wikidata":"https://www.wikidata.org/wiki/Q810247","display_name":"Baseline (sea)","level":2,"score":0.41110000014305115},{"id":"https://openalex.org/C2776175482","wikidata":"https://www.wikidata.org/wiki/Q1195816","display_name":"Transfer (computing)","level":2,"score":0.3808000087738037},{"id":"https://openalex.org/C150899416","wikidata":"https://www.wikidata.org/wiki/Q1820378","display_name":"Transfer of learning","level":2,"score":0.3783000111579895},{"id":"https://openalex.org/C2779808786","wikidata":"https://www.wikidata.org/wiki/Q6664603","display_name":"Locality","level":2,"score":0.3781000077724457},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.32659998536109924},{"id":"https://openalex.org/C126388530","wikidata":"https://www.wikidata.org/wiki/Q1131737","display_name":"Imitation","level":2,"score":0.30889999866485596},{"id":"https://openalex.org/C115961682","wikidata":"https://www.wikidata.org/wiki/Q860623","display_name":"Image (mathematics)","level":2,"score":0.3077000081539154},{"id":"https://openalex.org/C31972630","wikidata":"https://www.wikidata.org/wiki/Q844240","display_name":"Computer vision","level":1,"score":0.3066999912261963},{"id":"https://openalex.org/C105339364","wikidata":"https://www.wikidata.org/wiki/Q2297740","display_name":"Software deployment","level":2,"score":0.30329999327659607},{"id":"https://openalex.org/C126255220","wikidata":"https://www.wikidata.org/wiki/Q141495","display_name":"Mathematical optimization","level":1,"score":0.2915000021457672},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.26989999413490295},{"id":"https://openalex.org/C112972136","wikidata":"https://www.wikidata.org/wiki/Q7595718","display_name":"Stability (learning theory)","level":2,"score":0.2694000005722046},{"id":"https://openalex.org/C184898388","wikidata":"https://www.wikidata.org/wiki/Q1435712","display_name":"Pairwise comparison","level":2,"score":0.26579999923706055},{"id":"https://openalex.org/C22367795","wikidata":"https://www.wikidata.org/wiki/Q7625208","display_name":"Structured prediction","level":2,"score":0.25769999623298645},{"id":"https://openalex.org/C198352243","wikidata":"https://www.wikidata.org/wiki/Q37105","display_name":"Line (geometry)","level":2,"score":0.2531999945640564},{"id":"https://openalex.org/C110121322","wikidata":"https://www.wikidata.org/wiki/Q865811","display_name":"Distribution (mathematics)","level":2,"score":0.25040000677108765}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2609.32333","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.32333","pdf_url":"https://arxiv.org/pdf/2609.32333","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.2609.32333","doi":"https://doi.org/10.48550/arxiv.2609.32333","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2609.32333","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:2609.32333","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.32333","pdf_url":"https://arxiv.org/pdf/2609.32333","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":[{"display_name":"Quality education","score":0.4156000018119812,"id":"https://metadata.un.org/sdg/4"}],"study_designs":[],"awards":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7214778286.pdf","grobid_xml":"https://content.openalex.org/works/W7214778286.grobid-xml"},"referenced_works_count":6,"referenced_works":["https://openalex.org/W4391631327","https://openalex.org/W4414939355","https://openalex.org/W7161676639","https://openalex.org/W7162070791","https://openalex.org/W7162574072","https://openalex.org/W7201842969"],"related_works":["https://openalex.org/W7161676639","https://openalex.org/W7128783650","https://openalex.org/W7167628518","https://openalex.org/W7172386296","https://openalex.org/W7163681633","https://openalex.org/W7215041497","https://openalex.org/W7214728923","https://openalex.org/W7163551959","https://openalex.org/W7162070791","https://openalex.org/W7221574573"],"abstract_inverted_index":{"Regional-to-global":[0],"distillation":[1,28,151],"uses":[2],"crop-conditioned":[3,83],"guidance":[4],"to":[5,12,19,70],"improve":[6],"full-image":[7,22],"understanding.":[8],"The":[9,49],"challenge":[10],"is":[11,99],"effectively":[13],"transfer":[14],"the":[15,20,32,37,40,57,64,71,82,97,106,123,149,165],"teacher's":[16],"crop-based":[17],"advantage":[18],"student's":[21,33],"inference.":[23],"We":[24],"propose":[25],"progressive-view":[26],"on-policy":[27],"(PVD),":[29],"which":[30],"shifts":[31],"view":[34,65],"distribution":[35],"from":[36,119],"crop":[38,51],"toward":[39,122],"full":[41,58,72],"image":[42],"through":[43],"an":[44,139],"intermediate":[45],"aspect-preserving":[46],"padded":[47,50],"crop.":[48],"preserves":[52],"regional":[53],"content":[54],"while":[55],"matching":[56],"image's":[59],"visual-token":[60],"grid.":[61],"Across":[62,126],"stages,":[63],"mixture":[66],"assigns":[67],"increasing":[68],"probability":[69],"image.":[73],"A":[74,109],"lightweight":[75],"regional-advantage":[76],"weighting":[77],"reallocates":[78],"token-level":[79],"supervision":[80],"using":[81],"teacher-student":[84],"log-probability":[85],"gap.":[86],"Evaluated":[87],"under":[88],"each":[89],"sampled":[90],"input,":[91],"it":[92],"applies":[93],"mild":[94],"reweighting":[95],"when":[96,105],"gap":[98,107],"small":[100],"and":[101,132,156],"emphasizes":[102],"higher-gap":[103],"tokens":[104],"widens.":[108],"Jensen-Shannon":[110],"metric":[111],"decomposition":[112],"interprets":[113],"this":[114],"schedule":[115],"as":[116],"a":[117],"shift":[118],"matched-input":[120],"imitation":[121],"deployment":[124],"objective.":[125],"benchmarks":[127],"spanning":[128],"perception,":[129],"visual":[130],"mathematics":[131],"general":[133],"multimodal":[134],"question":[135],"answering,":[136],"PVD-full":[137],"reaches":[138],"average":[140],"accuracy":[141],"of":[142],"77.51":[143],"over":[144],"three":[145],"seeds,":[146],"improving":[147],"on":[148,157],"reward-free":[150,166],"baseline":[152],"by":[153,161],"2.01":[154],"points":[155],"its":[158],"reward-matched":[159],"variant":[160],"1.00":[162],"point.":[163],"In":[164],"setting,":[167],"PVD-distill":[168],"still":[169],"gains":[170],"1.16":[171],"points.":[172]},"counts_by_year":[],"updated_date":"2026-11-07T11:25:07.637180","created_date":"2026-09-30T00:00:00"}