{"id":"https://openalex.org/W7151768913","doi":"https://doi.org/10.48550/arxiv.2604.05868","title":"Understanding Performance Gap Between Parallel and Sequential Sampling in Large Reasoning Models","display_name":"Understanding Performance Gap Between Parallel and Sequential Sampling in Large Reasoning Models","publication_year":2026,"publication_date":"2026-04-07","ids":{"openalex":"https://openalex.org/W7151768913","doi":"https://doi.org/10.48550/arxiv.2604.05868"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2604.05868","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.05868","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","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.2604.05868","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5133175267","display_name":"Xiangming Gu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Gu, Xiangming","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5036322364","display_name":"Soham De","orcid":"https://orcid.org/0000-0001-7907-1335"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"De, Soham","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5003799439","display_name":"Larisa Markeeva","orcid":"https://orcid.org/0000-0002-7453-0827"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Markeeva, Larisa","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5008869927","display_name":"Petar Veli\u010dkovi\u0107","orcid":"https://orcid.org/0000-0002-2820-4692"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Veli\u010dkovi\u0107, Petar","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5043910056","display_name":"Razvan Pascanu","orcid":"https://orcid.org/0000-0002-5470-1238"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Pascanu, Razvan","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/T10028","display_name":"Topic Modeling","score":0.17659999430179596,"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"}},"topics":[{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.17659999430179596,"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/T12026","display_name":"Explainable Artificial Intelligence (XAI)","score":0.06270000338554382,"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/T12535","display_name":"Machine Learning and Data Classification","score":0.04230000078678131,"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/sampling","display_name":"Sampling (signal processing)","score":0.7035999894142151},{"id":"https://openalex.org/keywords/sequential-sampling","display_name":"Sequential sampling","score":0.6689000129699707},{"id":"https://openalex.org/keywords/sample","display_name":"Sample (material)","score":0.49810001254081726},{"id":"https://openalex.org/keywords/context","display_name":"Context (archaeology)","score":0.48829999566078186},{"id":"https://openalex.org/keywords/representation","display_name":"Representation (politics)","score":0.428600013256073},{"id":"https://openalex.org/keywords/sequential-estimation","display_name":"Sequential estimation","score":0.4036000072956085},{"id":"https://openalex.org/keywords/quality","display_name":"Quality (philosophy)","score":0.3847000002861023},{"id":"https://openalex.org/keywords/sampling-bias","display_name":"Sampling bias","score":0.3619000017642975}],"concepts":[{"id":"https://openalex.org/C140779682","wikidata":"https://www.wikidata.org/wiki/Q210868","display_name":"Sampling (signal processing)","level":3,"score":0.7035999894142151},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6862999796867371},{"id":"https://openalex.org/C2984391234","wikidata":"https://www.wikidata.org/wiki/Q195771","display_name":"Sequential sampling","level":3,"score":0.6689000129699707},{"id":"https://openalex.org/C198531522","wikidata":"https://www.wikidata.org/wiki/Q485146","display_name":"Sample (material)","level":2,"score":0.49810001254081726},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.48829999566078186},{"id":"https://openalex.org/C2779343474","wikidata":"https://www.wikidata.org/wiki/Q3109175","display_name":"Context (archaeology)","level":2,"score":0.48829999566078186},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.43479999899864197},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.4318000078201294},{"id":"https://openalex.org/C2776359362","wikidata":"https://www.wikidata.org/wiki/Q2145286","display_name":"Representation (politics)","level":3,"score":0.428600013256073},{"id":"https://openalex.org/C86426650","wikidata":"https://www.wikidata.org/wiki/Q7452504","display_name":"Sequential estimation","level":2,"score":0.4036000072956085},{"id":"https://openalex.org/C2779530757","wikidata":"https://www.wikidata.org/wiki/Q1207505","display_name":"Quality (philosophy)","level":2,"score":0.3847000002861023},{"id":"https://openalex.org/C75917345","wikidata":"https://www.wikidata.org/wiki/Q2725298","display_name":"Sampling bias","level":3,"score":0.3619000017642975},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.3569999933242798},{"id":"https://openalex.org/C106399304","wikidata":"https://www.wikidata.org/wiki/Q868123","display_name":"Snowball sampling","level":2,"score":0.35569998621940613},{"id":"https://openalex.org/C52740198","wikidata":"https://www.wikidata.org/wiki/Q1539564","display_name":"Importance sampling","level":3,"score":0.32179999351501465},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.3034000098705292},{"id":"https://openalex.org/C73602740","wikidata":"https://www.wikidata.org/wiki/Q7795822","display_name":"Thompson sampling","level":3,"score":0.298799991607666},{"id":"https://openalex.org/C80478641","wikidata":"https://www.wikidata.org/wiki/Q195771","display_name":"Sequential analysis","level":2,"score":0.2897000014781952},{"id":"https://openalex.org/C63584917","wikidata":"https://www.wikidata.org/wiki/Q333286","display_name":"Bounding overwatch","level":2,"score":0.2822999954223633},{"id":"https://openalex.org/C166052673","wikidata":"https://www.wikidata.org/wiki/Q83021","display_name":"Empirical evidence","level":2,"score":0.2752000093460083},{"id":"https://openalex.org/C158424031","wikidata":"https://www.wikidata.org/wiki/Q1191905","display_name":"Gibbs sampling","level":3,"score":0.27219998836517334},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.26460000872612},{"id":"https://openalex.org/C129848803","wikidata":"https://www.wikidata.org/wiki/Q2564360","display_name":"Sample size determination","level":2,"score":0.26100000739097595},{"id":"https://openalex.org/C167723999","wikidata":"https://www.wikidata.org/wiki/Q3773214","display_name":"Sampling distribution","level":2,"score":0.257099986076355}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2604.05868","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.05868","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2604.05868","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.05868","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/4","display_name":"Quality Education","score":0.8197452425956726}],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Large":[0],"Reasoning":[1],"Models":[2],"(LRMs)":[3],"have":[4,82],"shown":[5],"remarkable":[6],"performance":[7,172,198],"on":[8,95,131,137],"challenging":[9],"questions,":[10],"such":[11],"as":[12],"math":[13],"and":[14,49,63,141,149,153,159,187],"coding.":[15],"However,":[16],"to":[17,26,42,73,106,117,125,129,165,181],"obtain":[18],"a":[19,183],"high":[20],"quality":[21],"solution,":[22],"one":[23,193],"may":[24],"need":[25],"sample":[27],"more":[28,44,83],"than":[29],"once.":[30],"In":[31,52,174],"principal,":[32],"there":[33],"are":[34],"two":[35,59],"sampling":[36,48,71,76,103,112,123],"strategies":[37],"that":[38,69,156,190],"can":[39],"be":[40,166],"composed":[41],"form":[43],"complex":[45],"processes:":[46],"sequential":[47,75,111,122],"parallel":[50,70,102],"sampling.":[51],"this":[53,99,191],"paper,":[54],"we":[55,91,188],"first":[56],"compare":[57],"these":[58],"approaches":[60],"with":[61,66],"rigor,":[62],"observe,":[64],"aligned":[65],"previous":[67,132],"works,":[68],"seems":[72,180],"outperform":[74],"even":[77],"though":[78],"the":[79,88,96,107,157,167,171,176,197],"latter":[80],"should":[81],"representation":[84],"power.":[85],"To":[86],"understand":[87],"underline":[89],"reasons,":[90],"make":[92],"three":[93],"hypothesis":[94],"reason":[97],"behind":[98,170],"behavior:":[100],"(i)":[101],"outperforms":[104],"due":[105,128],"aggregator":[108],"operator;":[109],"(ii)":[110],"is":[113,192],"harmed":[114],"by":[115],"needing":[116],"use":[118],"longer":[119],"contexts;":[120],"(iii)":[121],"leads":[124],"less":[126],"exploration":[127,179],"conditioning":[130],"answers.":[133],"The":[134],"empirical":[135],"evidence":[136],"various":[138],"model":[139],"families":[140],"sizes":[142],"(Qwen3,":[143],"DeepSeek-R1":[144],"distilled":[145],"models,":[146],"Gemini":[147],"2.5)":[148],"question":[150],"domains":[151],"(math":[152],"coding)":[154],"suggests":[155],"aggregation":[158],"context":[160],"length":[161],"do":[162],"not":[163],"seem":[164],"main":[168,194],"culprit":[169],"gap.":[173,199],"contrast,":[175],"lack":[177],"of":[178],"play":[182],"considerably":[184],"larger":[185],"role,":[186],"argue":[189],"cause":[195],"for":[196]},"counts_by_year":[],"updated_date":"2026-07-28T07:46:37.118299","created_date":"2026-04-09T00:00:00"}