{"id":"https://openalex.org/W7202128551","doi":"https://doi.org/10.48550/arxiv.2608.06794","title":"PAST: Prompt-Adaptive Sampling Termination for Efficient Diffusion Model","display_name":"PAST: Prompt-Adaptive Sampling Termination for Efficient Diffusion Model","publication_year":2026,"publication_date":"2026-08-07","ids":{"openalex":"https://openalex.org/W7202128551","doi":"https://doi.org/10.48550/arxiv.2608.06794"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2608.06794","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.06794","pdf_url":"https://arxiv.org/pdf/2608.06794","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":"public-domain","license_id":"https://openalex.org/licenses/public-domain","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/2608.06794","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5089763177","display_name":"Renye Yan","orcid":"https://orcid.org/0009-0005-9057-9603","observed_orcids":["https://orcid.org/0009-0005-9057-9603"]},"institutions":[{"id":"https://openalex.org/I20231570","display_name":"Peking University","ror":"https://ror.org/02v51f717","country_code":"CN","type":"education","lineage":["https://openalex.org/I20231570"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yan, Renye","raw_affiliation_strings":["Peking University Beijing , China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Peking University Beijing , China","institution_ids":["https://openalex.org/I20231570"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5041019639","display_name":"Jikang Cheng","orcid":"https://orcid.org/0000-0001-6549-6148","observed_orcids":["https://orcid.org/0000-0001-6549-6148"]},"institutions":[{"id":"https://openalex.org/I20231570","display_name":"Peking University","ror":"https://ror.org/02v51f717","country_code":"CN","type":"education","lineage":["https://openalex.org/I20231570"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Cheng, Jikang","raw_affiliation_strings":["Peking University Beijing , China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Peking University Beijing , China","institution_ids":["https://openalex.org/I20231570"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5146287553","display_name":"You Wu","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I881766915","display_name":"Nanjing University","ror":"https://ror.org/01rxvg760","country_code":"CN","type":"education","lineage":["https://openalex.org/I881766915"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Wu, You","raw_affiliation_strings":["Nanjing University Nanjing , China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Nanjing University Nanjing , China","institution_ids":["https://openalex.org/I881766915"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5019479187","display_name":"Wei Peng","orcid":"https://orcid.org/0000-0002-2892-5764","observed_orcids":["https://orcid.org/0000-0002-2892-5764"]},"institutions":[{"id":"https://openalex.org/I97018004","display_name":"Stanford University","ror":"https://ror.org/00f54p054","country_code":"US","type":"education","lineage":["https://openalex.org/I97018004"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Peng, Wei","raw_affiliation_strings":["Stanford University Stanford , California , USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Stanford University Stanford , California , USA","institution_ids":["https://openalex.org/I97018004"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5146474268","display_name":"Zongwei Wang","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I20231570","display_name":"Peking University","ror":"https://ror.org/02v51f717","country_code":"CN","type":"education","lineage":["https://openalex.org/I20231570"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Wang, Zongwei","raw_affiliation_strings":["Peking University Beijing , China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Peking University Beijing , China","institution_ids":["https://openalex.org/I20231570"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5146362624","display_name":"Ling Liang","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I20231570","display_name":"Peking University","ror":"https://ror.org/02v51f717","country_code":"CN","type":"education","lineage":["https://openalex.org/I20231570"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Liang, Ling","raw_affiliation_strings":["Peking University Beijing , China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Peking University Beijing , China","institution_ids":["https://openalex.org/I20231570"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5136412184","display_name":"Yimao Cai","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I20231570","display_name":"Peking University","ror":"https://ror.org/02v51f717","country_code":"CN","type":"education","lineage":["https://openalex.org/I20231570"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Cai, Yimao","raw_affiliation_strings":["Peking University Beijing , China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Peking University Beijing , China","institution_ids":["https://openalex.org/I20231570"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":3,"corresponding_author_ids":[],"corresponding_institution_ids":[],"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/T10775","display_name":"Generative Adversarial Networks and Image Synthesis","score":0.9994999766349792,"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":0.9994999766349792,"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/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.00019999999494757503,"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/T11714","display_name":"Multimodal Machine Learning Applications","score":9.999999747378752e-05,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/diffusion-models","display_name":"diffusion models","score":0.8560000061988831},{"id":"https://openalex.org/keywords/text-to-image-generation","display_name":"text-to-image generation","score":0.8029999732971191},{"id":"https://openalex.org/keywords/reinforcement-learning-fine-tuning","display_name":"reinforcement learning fine-tuning","score":0.878000020980835},{"id":"https://openalex.org/keywords/intrinsic-rewards","display_name":"intrinsic rewards","score":0.6330000162124634},{"id":"https://openalex.org/keywords/semantic-alignment","display_name":"semantic alignment","score":0.6460000276565552},{"id":"https://openalex.org/keywords/computational-efficiency","display_name":"computational efficiency","score":0.7289999723434448}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6972000002861023},{"id":"https://openalex.org/C99498987","wikidata":"https://www.wikidata.org/wiki/Q2210247","display_name":"Noise (video)","level":3,"score":0.6410999894142151},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.5738000273704529},{"id":"https://openalex.org/C163294075","wikidata":"https://www.wikidata.org/wiki/Q581861","display_name":"Noise reduction","level":2,"score":0.5601000189781189},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4708999991416931},{"id":"https://openalex.org/C2779530757","wikidata":"https://www.wikidata.org/wiki/Q1207505","display_name":"Quality (philosophy)","level":2,"score":0.4641000032424927},{"id":"https://openalex.org/C140779682","wikidata":"https://www.wikidata.org/wiki/Q210868","display_name":"Sampling (signal processing)","level":3,"score":0.4478999972343445},{"id":"https://openalex.org/C2780980858","wikidata":"https://www.wikidata.org/wiki/Q110022","display_name":"Dual (grammatical number)","level":2,"score":0.4442000091075897},{"id":"https://openalex.org/C2781395549","wikidata":"https://www.wikidata.org/wiki/Q4680762","display_name":"Adaptive sampling","level":3,"score":0.4235999882221222},{"id":"https://openalex.org/C111335779","wikidata":"https://www.wikidata.org/wiki/Q3454686","display_name":"Reduction (mathematics)","level":2,"score":0.4020000100135803},{"id":"https://openalex.org/C155512373","wikidata":"https://www.wikidata.org/wiki/Q287450","display_name":"Residual","level":2,"score":0.3937999904155731},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.3765999972820282},{"id":"https://openalex.org/C126255220","wikidata":"https://www.wikidata.org/wiki/Q141495","display_name":"Mathematical optimization","level":1,"score":0.3562999963760376},{"id":"https://openalex.org/C69357855","wikidata":"https://www.wikidata.org/wiki/Q163214","display_name":"Diffusion","level":2,"score":0.3375999927520752},{"id":"https://openalex.org/C52740198","wikidata":"https://www.wikidata.org/wiki/Q1539564","display_name":"Importance sampling","level":3,"score":0.32760000228881836},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.30489999055862427},{"id":"https://openalex.org/C2780009758","wikidata":"https://www.wikidata.org/wiki/Q6804172","display_name":"Measure (data warehouse)","level":2,"score":0.2768000066280365},{"id":"https://openalex.org/C2780598303","wikidata":"https://www.wikidata.org/wiki/Q65921492","display_name":"Flexibility (engineering)","level":2,"score":0.2669000029563904},{"id":"https://openalex.org/C55020928","wikidata":"https://www.wikidata.org/wiki/Q3813865","display_name":"Image quality","level":3,"score":0.257999986410141},{"id":"https://openalex.org/C2781249084","wikidata":"https://www.wikidata.org/wiki/Q908656","display_name":"Preference","level":2,"score":0.25609999895095825}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2608.06794","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.06794","pdf_url":"https://arxiv.org/pdf/2608.06794","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":"public-domain","license_id":"https://openalex.org/licenses/public-domain","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2608.06794","doi":"https://doi.org/10.48550/arxiv.2608.06794","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2608.06794","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":"public-domain","license_id":"https://openalex.org/licenses/public-domain","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2608.06794","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.06794","pdf_url":"https://arxiv.org/pdf/2608.06794","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":"public-domain","license_id":"https://openalex.org/licenses/public-domain","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7202128551.pdf","grobid_xml":"https://content.openalex.org/works/W7202128551.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W7161586305","https://openalex.org/W4416843616","https://openalex.org/W4413147466","https://openalex.org/W4412494096","https://openalex.org/W4402727620","https://openalex.org/W4378464471","https://openalex.org/W4407759100","https://openalex.org/W7138161599","https://openalex.org/W4387294554","https://openalex.org/W4416110871"],"abstract_inverted_index":{"While":[0],"diffusion":[1],"models":[2],"have":[3],"made":[4],"significant":[5],"progress":[6,58],"in":[7],"text-to-image":[8],"tasks,":[9],"they":[10],"still":[11],"exhibit":[12],"limitations":[13],"when":[14],"directly":[15],"optimizing":[16],"downstream":[17],"objectives.":[18],"Although":[19],"Reinforcement":[20],"Learning":[21],"(RL)":[22],"enables":[23,124],"targeted":[24],"optimization,":[25],"existing":[26,180],"methods":[27,183],"are":[28],"generally":[29],"constrained":[30],"by":[31,54,184,193],"low-efficiency":[32],"fine-tuning":[33,182],"and":[34,59,75,103,109,134,161,169],"sparse":[35,72],"rewards.":[36,96],"To":[37],"address":[38],"these":[39],"challenges,":[40],"we":[41,63,147],"propose":[42],"PAST,":[43],"which":[44],"provides":[45],"differentiated":[46],"rewards":[47,74,163],"while":[48,188],"adaptively":[49,120],"regulating":[50],"training":[51],"episode":[52,128],"length":[53],"jointly":[55],"perceiving":[56],"denoising":[57,101],"prompt":[60,110,132],"difficulty.":[61],"Specifically,":[62,154],"design":[64],"an":[65],"intrinsic":[66,95,162],"reward":[67],"paradigm":[68],"to":[69,79,186,195],"compensate":[70],"for":[71,94],"extrinsic":[73,160],"guide":[76],"the":[77,118,135,142,159,167],"model":[78],"explore":[80],"paths":[81],"that":[82,174],"diverge":[83],"more":[84],"efficiently":[85],"from":[86],"noise":[87,145],"patterns.":[88],"We":[89],"further":[90],"provide":[91],"theoretical":[92],"justification":[93],"Then,":[97],"PAST":[98,175],"dynamically":[99],"monitors":[100],"completion":[102],"semantic":[104],"alignment":[105],"between":[106],"image":[107],"structures":[108],"semantics.":[111],"When":[112],"both":[113],"metrics":[114],"satisfy":[115],"generation":[116,137],"requirements,":[117],"system":[119],"terminates":[121],"training.":[122],"This":[123],"appropriate":[125],"allocation":[126],"of":[127,179],"lengths":[129],"based":[130,140],"on":[131,141],"difficulty":[133],"current":[136],"process.":[138],"Finally,":[139],"predicted":[143],"residual":[144],"level,":[146],"establish":[148],"a":[149],"dual":[150,199],"adaptive":[151,200],"coordination":[152],"mechanism.":[153,202],"it":[155],"not":[156],"only":[157],"balances":[158,166],"but":[164],"also":[165],"exploration":[168],"convergence.":[170],"Experimental":[171],"results":[172],"demonstrate":[173],"enhances":[176],"computational":[177],"efficiency":[178],"RL":[181],"up":[185,194],"66.7%,":[187],"improving":[189],"preference":[190],"optimization":[191],"quality":[192],"29.5%":[196],"through":[197],"its":[198],"regulation":[201]},"counts_by_year":[],"updated_date":"2026-11-04T10:54:05.112534","created_date":"2026-08-11T00:00:00"}