{"id":"https://openalex.org/W7162751592","doi":"https://doi.org/10.48550/arxiv.2605.30327","title":"Reasoning with Sampling: Cutting at Decision Points","display_name":"Reasoning with Sampling: Cutting at Decision Points","publication_year":2026,"publication_date":"2026-05-28","ids":{"openalex":"https://openalex.org/W7162751592","doi":"https://doi.org/10.48550/arxiv.2605.30327"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2605.30327","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.30327","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.2605.30327","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5137315609","display_name":"Felix Zhou","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhou, Felix","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5065512381","display_name":"Anay Mehrotra","orcid":"https://orcid.org/0000-0002-8566-5452"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Mehrotra, Anay","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5137351908","display_name":"Quanquan C. Liu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liu, Quanquan C.","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/T10181","display_name":"Natural Language Processing Techniques","score":0.4675000011920929,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.4675000011920929,"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/T10028","display_name":"Topic Modeling","score":0.2987000048160553,"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":0.04349999874830246,"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/succinctness","display_name":"Succinctness","score":0.5389999747276306},{"id":"https://openalex.org/keywords/entropy","display_name":"Entropy (arrow of time)","score":0.5060999989509583},{"id":"https://openalex.org/keywords/stylized-fact","display_name":"Stylized fact","score":0.4772999882698059},{"id":"https://openalex.org/keywords/trace","display_name":"TRACE (psycholinguistics)","score":0.40119999647140503},{"id":"https://openalex.org/keywords/base","display_name":"Base (topology)","score":0.3528999984264374},{"id":"https://openalex.org/keywords/decision-tree","display_name":"Decision tree","score":0.3479999899864197},{"id":"https://openalex.org/keywords/proxy","display_name":"Proxy (statistics)","score":0.34630000591278076},{"id":"https://openalex.org/keywords/knowledge-base","display_name":"Knowledge base","score":0.3409000039100647}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6574000120162964},{"id":"https://openalex.org/C2776493592","wikidata":"https://www.wikidata.org/wiki/Q5158717","display_name":"Succinctness","level":2,"score":0.5389999747276306},{"id":"https://openalex.org/C106301342","wikidata":"https://www.wikidata.org/wiki/Q4117933","display_name":"Entropy (arrow of time)","level":2,"score":0.5060999989509583},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.48260000348091125},{"id":"https://openalex.org/C38935604","wikidata":"https://www.wikidata.org/wiki/Q4330363","display_name":"Stylized fact","level":2,"score":0.4772999882698059},{"id":"https://openalex.org/C75291252","wikidata":"https://www.wikidata.org/wiki/Q1315756","display_name":"TRACE (psycholinguistics)","level":2,"score":0.40119999647140503},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.36230000853538513},{"id":"https://openalex.org/C42058472","wikidata":"https://www.wikidata.org/wiki/Q810214","display_name":"Base (topology)","level":2,"score":0.3528999984264374},{"id":"https://openalex.org/C84525736","wikidata":"https://www.wikidata.org/wiki/Q831366","display_name":"Decision tree","level":2,"score":0.3479999899864197},{"id":"https://openalex.org/C2780148112","wikidata":"https://www.wikidata.org/wiki/Q1432581","display_name":"Proxy (statistics)","level":2,"score":0.34630000591278076},{"id":"https://openalex.org/C4554734","wikidata":"https://www.wikidata.org/wiki/Q593744","display_name":"Knowledge base","level":2,"score":0.3409000039100647},{"id":"https://openalex.org/C2986080485","wikidata":"https://www.wikidata.org/wiki/Q1331926","display_name":"Decision maker","level":2,"score":0.3222000002861023},{"id":"https://openalex.org/C198082294","wikidata":"https://www.wikidata.org/wiki/Q3399648","display_name":"Position (finance)","level":2,"score":0.31139999628067017},{"id":"https://openalex.org/C2985612853","wikidata":"https://www.wikidata.org/wiki/Q185816","display_name":"Analogical reasoning","level":3,"score":0.29249998927116394},{"id":"https://openalex.org/C140779682","wikidata":"https://www.wikidata.org/wiki/Q210868","display_name":"Sampling (signal processing)","level":3,"score":0.2865000069141388},{"id":"https://openalex.org/C2777508537","wikidata":"https://www.wikidata.org/wiki/Q7936620","display_name":"Visual reasoning","level":2,"score":0.28540000319480896},{"id":"https://openalex.org/C83725634","wikidata":"https://www.wikidata.org/wiki/Q7268699","display_name":"Qualitative reasoning","level":2,"score":0.2800999879837036},{"id":"https://openalex.org/C9679016","wikidata":"https://www.wikidata.org/wiki/Q1417473","display_name":"Principle of maximum entropy","level":2,"score":0.2694999873638153},{"id":"https://openalex.org/C115988155","wikidata":"https://www.wikidata.org/wiki/Q3262192","display_name":"Decision problem","level":2,"score":0.26910001039505005},{"id":"https://openalex.org/C37335422","wikidata":"https://www.wikidata.org/wiki/Q6888134","display_name":"Model-based reasoning","level":3,"score":0.26499998569488525},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.2565999925136566},{"id":"https://openalex.org/C156201811","wikidata":"https://www.wikidata.org/wiki/Q5418360","display_name":"Evidential reasoning approach","level":4,"score":0.25619998574256897},{"id":"https://openalex.org/C195818886","wikidata":"https://www.wikidata.org/wiki/Q5421724","display_name":"Expressive power","level":2,"score":0.25529998540878296},{"id":"https://openalex.org/C88626702","wikidata":"https://www.wikidata.org/wiki/Q1128903","display_name":"Continuation","level":2,"score":0.25429999828338623}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2605.30327","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.30327","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.2605.30327","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.30327","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":[{"id":"https://metadata.un.org/sdg/16","display_name":"Peace, Justice and strong institutions","score":0.7212287783622742}],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Frontier":[0],"reasoning":[1,37,79,95,109],"models":[2,9],"are":[3,175],"produced":[4],"by":[5,18],"posttraining":[6],"base":[7,28,152],"language":[8],"with":[10,196,206],"reinforcement":[11],"learning.":[12],"Recent":[13],"work":[14],"has":[15],"challenged":[16],"this":[17,47],"showing":[19],"that":[20,105,128,149,172,190],"sampling":[21,52],"from":[22,53,104,166],"a":[23,31,89,113,129,157,176,184,202],"sharpened":[24],"version":[25],"of":[26,71,120,187,199,209],"the":[27,54,63,72,93,102,118,151,197,207],"model's":[29,153],"distribution,":[30,34,65],"so-called":[32],"power":[33,55,64],"elicits":[35],"comparable":[36],"without":[38],"additional":[39],"training,":[40],"curated":[41],"datasets,":[42],"or":[43,123],"verifiers.":[44],"However,":[45,108],"making":[46],"method":[48,224],"practical":[49],"requires":[50],"efficiently":[51],"distribution.":[56],"A":[57],"sampler":[58],"needs":[59],"to":[60,62,134,159],"\"mix\"":[61],"which":[66,211],"necessitates":[67],"moving":[68],"between":[69],"modes":[70],"target":[73],"distribution;":[74],"intuitively,":[75],"e.g.,":[76],"trying":[77],"different":[78],"strategies.":[80],"The":[81],"samplers":[82],"proposed":[83],"in":[84,92,183,201],"prior":[85],"works":[86],"repeatedly":[87],"select":[88],"\"cut\"":[90],"position":[91,106],"current":[94],"trace":[96,203],"uniformly":[97,130],"at":[98],"random":[99],"and":[100,125,164,221,229],"resample":[101,165],"suffix":[103],"onward.":[107],"traces":[110],"typically":[111],"contain":[112],"few":[114],"consequential":[115],"decisions":[116,200],"(e.g.,":[117],"choice":[119],"proof":[121],"strategy":[122],"algorithm),":[124],"we":[126],"observe":[127],"chosen":[131],"cut":[132],"tends":[133],"rewrite":[135],"local":[136],"details":[137],"rather":[138,204],"than":[139,205],"revisit":[140],"decision":[141,162,180],"points.":[142],"We":[143,169],"introduce":[144],"an":[145],"algorithm":[146],"(Entropy-Cut":[147],"Metropolis-Hastings)":[148],"uses":[150],"next-token":[154],"entropy":[155,173],"as":[156],"proxy":[158,178],"identify":[160],"key":[161],"points":[163,181],"those":[167],"positions.":[168],"empirically":[170],"verify":[171],"jumps":[174],"useful":[177],"for":[179],"and,":[182],"stylized":[185],"model":[186],"reasoning,":[188],"prove":[189],"our":[191,223],"method's":[192],"mixing":[193],"time":[194],"scales":[195],"number":[198,208],"tokens,":[210],"can":[212],"be":[213],"much":[214],"larger.":[215],"Across":[216],"MATH500,":[217],"HumanEval,":[218],"GPQA":[219],"Diamond,":[220],"AIME26,":[222],"consistently":[225],"improves":[226],"over":[227],"baselines":[228],"RL-trained":[230],"models.":[231]},"counts_by_year":[],"updated_date":"2026-07-28T07:46:37.118299","created_date":"2026-05-30T00:00:00"}