{"id":"https://openalex.org/W7163050691","doi":"https://doi.org/10.48550/arxiv.2605.30991","title":"Parallel Tempering Initial Sampling in Inference-Time Reward Alignment","display_name":"Parallel Tempering Initial Sampling in Inference-Time Reward Alignment","publication_year":2026,"publication_date":"2026-05-29","ids":{"openalex":"https://openalex.org/W7163050691","doi":"https://doi.org/10.48550/arxiv.2605.30991"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2605.30991","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.30991","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,"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.2605.30991","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5137608445","display_name":"Myeongjun Oh","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Oh, Myeongjun","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137573029","display_name":"Gwangho Kim","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Kim, Gwangho","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5101790501","display_name":"Sungyoon Lee","orcid":"https://orcid.org/0000-0002-6097-0565"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lee, Sungyoon","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":0.9344000220298767,"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":0.9344000220298767,"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":0.011500000022351742,"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/T13702","display_name":"Machine Learning in Healthcare","score":0.007799999788403511,"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/initialization","display_name":"Initialization","score":0.8159000277519226},{"id":"https://openalex.org/keywords/importance-sampling","display_name":"Importance sampling","score":0.6352999806404114},{"id":"https://openalex.org/keywords/sampling","display_name":"Sampling (signal processing)","score":0.6294000148773193},{"id":"https://openalex.org/keywords/task","display_name":"Task (project management)","score":0.5979999899864197},{"id":"https://openalex.org/keywords/parallel-tempering","display_name":"Parallel tempering","score":0.5468000173568726},{"id":"https://openalex.org/keywords/monte-carlo-method","display_name":"Monte Carlo method","score":0.4415999948978424},{"id":"https://openalex.org/keywords/mechanism","display_name":"Mechanism (biology)","score":0.35600000619888306},{"id":"https://openalex.org/keywords/sample","display_name":"Sample (material)","score":0.3334999978542328}],"concepts":[{"id":"https://openalex.org/C114466953","wikidata":"https://www.wikidata.org/wiki/Q6034165","display_name":"Initialization","level":2,"score":0.8159000277519226},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6747999787330627},{"id":"https://openalex.org/C52740198","wikidata":"https://www.wikidata.org/wiki/Q1539564","display_name":"Importance sampling","level":3,"score":0.6352999806404114},{"id":"https://openalex.org/C140779682","wikidata":"https://www.wikidata.org/wiki/Q210868","display_name":"Sampling (signal processing)","level":3,"score":0.6294000148773193},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.5979999899864197},{"id":"https://openalex.org/C187653413","wikidata":"https://www.wikidata.org/wiki/Q7135015","display_name":"Parallel tempering","level":5,"score":0.5468000173568726},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4803999960422516},{"id":"https://openalex.org/C19499675","wikidata":"https://www.wikidata.org/wiki/Q232207","display_name":"Monte Carlo method","level":2,"score":0.4415999948978424},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.4327000081539154},{"id":"https://openalex.org/C89611455","wikidata":"https://www.wikidata.org/wiki/Q6804646","display_name":"Mechanism (biology)","level":2,"score":0.35600000619888306},{"id":"https://openalex.org/C198531522","wikidata":"https://www.wikidata.org/wiki/Q485146","display_name":"Sample (material)","level":2,"score":0.3334999978542328},{"id":"https://openalex.org/C39890363","wikidata":"https://www.wikidata.org/wiki/Q36108","display_name":"Generative grammar","level":2,"score":0.3109000027179718},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.30790001153945923},{"id":"https://openalex.org/C167966045","wikidata":"https://www.wikidata.org/wiki/Q5532625","display_name":"Generative model","level":3,"score":0.3077999949455261},{"id":"https://openalex.org/C111350023","wikidata":"https://www.wikidata.org/wiki/Q1191869","display_name":"Markov chain Monte Carlo","level":3,"score":0.30489999055862427},{"id":"https://openalex.org/C13662910","wikidata":"https://www.wikidata.org/wiki/Q193139","display_name":"Trajectory","level":2,"score":0.28769999742507935},{"id":"https://openalex.org/C158424031","wikidata":"https://www.wikidata.org/wiki/Q1191905","display_name":"Gibbs sampling","level":3,"score":0.271699994802475},{"id":"https://openalex.org/C52421305","wikidata":"https://www.wikidata.org/wiki/Q1151499","display_name":"Particle filter","level":3,"score":0.2653999924659729},{"id":"https://openalex.org/C187192777","wikidata":"https://www.wikidata.org/wiki/Q381699","display_name":"Rejection sampling","level":5,"score":0.26330000162124634},{"id":"https://openalex.org/C175154964","wikidata":"https://www.wikidata.org/wiki/Q380077","display_name":"Task analysis","level":3,"score":0.25769999623298645},{"id":"https://openalex.org/C73602740","wikidata":"https://www.wikidata.org/wiki/Q7795822","display_name":"Thompson sampling","level":3,"score":0.25099998712539673}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2605.30991","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.30991","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,"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.2605.30991","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.30991","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,"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":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Inference-time":[0],"reward":[1,62,87,103,134],"alignment":[2,175],"steers":[3],"pretrained":[4],"diffusion":[5],"and":[6,33,124,165],"flow-based":[7],"generative":[8],"models":[9],"to":[10,79,160],"satisfy":[11],"user-specified":[12],"rewards":[13],"without":[14],"retraining.":[15],"Recently,":[16],"Sequential":[17],"Monte":[18],"Carlo":[19],"(SMC)":[20],"has":[21],"emerged":[22],"as":[23,85],"a":[24,54,105,119],"powerful":[25],"framework":[26],"for":[27,101],"this":[28,145],"task":[29],"by":[30],"iteratively":[31],"filtering":[32],"propagating":[34],"multiple":[35,111],"particles.":[36],"However,":[37],"we":[38,68,96],"show":[39,69,168],"that":[40,70,109,144,156,169],"standard":[41,55],"SMC-based":[42],"methods":[43],"often":[44,90],"suffer":[45],"from":[46,53],"poor":[47],"performance":[48],"because":[49],"they":[50],"initialize":[51],"particles":[52],"prior,":[56],"whereas":[57],"high-reward":[58,154],"regions":[59,155],"in":[60,82,174],"complex":[61,86,179],"landscapes":[63,88],"are":[64,89,157],"extremely":[65],"rare.":[66],"Further,":[67],"even":[71],"recent":[72],"reward-aware":[73],"initial":[74],"sampling":[75,112],"approaches":[76],"remain":[77],"vulnerable":[78],"getting":[80],"trapped":[81],"local":[83],"modes,":[84],"multi-modal.":[91],"To":[92],"overcome":[93],"these":[94],"limitations,":[95],"propose":[97],"PATHS":[98,117,170],"(PArallel":[99],"Tempering":[100],"High-complexity":[102],"Sampling),":[104],"novel":[106],"initialization":[107],"method":[108],"couples":[110],"chains":[113,123],"through":[114],"parallel":[115],"tempering.":[116],"maintains":[118],"ladder":[120],"of":[121,152],"reward-tempered":[122],"periodically":[125],"performs":[126],"Metropolis":[127],"swaps,":[128],"enabling":[129],"efficient":[130],"exploration":[131,151],"across":[132],"flattened":[133],"landscapes,":[135],"thereby":[136],"mitigating":[137],"the":[138,149],"mode-trapping":[139],"issues.":[140],"Our":[141],"analysis":[142],"reveals":[143],"mechanism":[146],"substantially":[147],"enhances":[148],"finite-budget":[150],"rare,":[153],"typically":[158],"challenging":[159],"sample.":[161],"Experiments":[162],"on":[163,178],"layout-to-image":[164],"quantity-aware":[166],"generation":[167],"achieves":[171],"consistent":[172],"gains":[173],"quality,":[176],"particularly":[177],"prompts.":[180]},"counts_by_year":[],"updated_date":"2026-07-28T07:46:37.118299","created_date":"2026-06-02T00:00:00"}