{"id":"https://openalex.org/W7135011161","doi":"https://doi.org/10.48550/arxiv.2603.10848","title":"$V_{0.5}$: Generalist Value Model as a Prior for Sparse RL Rollouts","display_name":"$V_{0.5}$: Generalist Value Model as a Prior for Sparse RL Rollouts","publication_year":2026,"publication_date":"2026-03-11","ids":{"openalex":"https://openalex.org/W7135011161","doi":"https://doi.org/10.48550/arxiv.2603.10848"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2603.10848","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.10848","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"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.2603.10848","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5128818014","display_name":"Yi-Kai Zhang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhang, Yi-Kai","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5128859866","display_name":"Yueqing Sun","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Sun, Yueqing","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5128833321","display_name":"Hongyan Hao","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Hao, Hongyan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5128902017","display_name":"Qi Gu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Gu, Qi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5128886489","display_name":"Xunliang Cai","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Cai, Xunliang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5128861638","display_name":"De-Chuan Zhan","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhan, De-Chuan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5128809448","display_name":"Han-Jia Ye","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ye, Han-Jia","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/T10462","display_name":"Reinforcement Learning in Robotics","score":0.8508999943733215,"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/T10462","display_name":"Reinforcement Learning in Robotics","score":0.8508999943733215,"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":0.017899999395012856,"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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.01679999940097332,"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/baseline","display_name":"Baseline (sea)","score":0.7260000109672546},{"id":"https://openalex.org/keywords/reinforcement-learning","display_name":"Reinforcement learning","score":0.5813999772071838},{"id":"https://openalex.org/keywords/verifiable-secret-sharing","display_name":"Verifiable secret sharing","score":0.5353999733924866},{"id":"https://openalex.org/keywords/variance","display_name":"Variance (accounting)","score":0.49709999561309814},{"id":"https://openalex.org/keywords/value","display_name":"Value (mathematics)","score":0.48330000042915344},{"id":"https://openalex.org/keywords/reliability","display_name":"Reliability (semiconductor)","score":0.4442000091075897},{"id":"https://openalex.org/keywords/convergence","display_name":"Convergence (economics)","score":0.42739999294281006},{"id":"https://openalex.org/keywords/encoding","display_name":"Encoding (memory)","score":0.38920000195503235}],"concepts":[{"id":"https://openalex.org/C12725497","wikidata":"https://www.wikidata.org/wiki/Q810247","display_name":"Baseline (sea)","level":2,"score":0.7260000109672546},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6101999878883362},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.5813999772071838},{"id":"https://openalex.org/C85847156","wikidata":"https://www.wikidata.org/wiki/Q59015987","display_name":"Verifiable secret sharing","level":3,"score":0.5353999733924866},{"id":"https://openalex.org/C196083921","wikidata":"https://www.wikidata.org/wiki/Q7915758","display_name":"Variance (accounting)","level":2,"score":0.49709999561309814},{"id":"https://openalex.org/C2776291640","wikidata":"https://www.wikidata.org/wiki/Q2912517","display_name":"Value (mathematics)","level":2,"score":0.48330000042915344},{"id":"https://openalex.org/C126255220","wikidata":"https://www.wikidata.org/wiki/Q141495","display_name":"Mathematical optimization","level":1,"score":0.45159998536109924},{"id":"https://openalex.org/C43214815","wikidata":"https://www.wikidata.org/wiki/Q7310987","display_name":"Reliability (semiconductor)","level":3,"score":0.4442000091075897},{"id":"https://openalex.org/C2777303404","wikidata":"https://www.wikidata.org/wiki/Q759757","display_name":"Convergence (economics)","level":2,"score":0.42739999294281006},{"id":"https://openalex.org/C125411270","wikidata":"https://www.wikidata.org/wiki/Q18653","display_name":"Encoding (memory)","level":2,"score":0.38920000195503235},{"id":"https://openalex.org/C57869625","wikidata":"https://www.wikidata.org/wiki/Q1783502","display_name":"Rate of convergence","level":3,"score":0.34040001034736633},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.3287999927997589},{"id":"https://openalex.org/C52740198","wikidata":"https://www.wikidata.org/wiki/Q1539564","display_name":"Importance sampling","level":3,"score":0.3093000054359436},{"id":"https://openalex.org/C8505890","wikidata":"https://www.wikidata.org/wiki/Q605095","display_name":"Budget constraint","level":2,"score":0.3012999892234802},{"id":"https://openalex.org/C140779682","wikidata":"https://www.wikidata.org/wiki/Q210868","display_name":"Sampling (signal processing)","level":3,"score":0.29510000348091125},{"id":"https://openalex.org/C114289077","wikidata":"https://www.wikidata.org/wiki/Q3284399","display_name":"Statistical model","level":2,"score":0.29420000314712524},{"id":"https://openalex.org/C120936955","wikidata":"https://www.wikidata.org/wiki/Q2155640","display_name":"Empirical research","level":2,"score":0.2847000062465668},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.2802000045776367},{"id":"https://openalex.org/C139945424","wikidata":"https://www.wikidata.org/wiki/Q1940696","display_name":"Mean squared error","level":2,"score":0.2711000144481659},{"id":"https://openalex.org/C149782125","wikidata":"https://www.wikidata.org/wiki/Q160039","display_name":"Econometrics","level":1,"score":0.26829999685287476},{"id":"https://openalex.org/C14646407","wikidata":"https://www.wikidata.org/wiki/Q1430750","display_name":"Bellman equation","level":2,"score":0.26460000872612},{"id":"https://openalex.org/C49937458","wikidata":"https://www.wikidata.org/wiki/Q2599292","display_name":"Probabilistic logic","level":2,"score":0.2632000148296356},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.257999986410141},{"id":"https://openalex.org/C141042865","wikidata":"https://www.wikidata.org/wiki/Q200125","display_name":"Expected value","level":2,"score":0.2578999996185303},{"id":"https://openalex.org/C61797465","wikidata":"https://www.wikidata.org/wiki/Q1188986","display_name":"Term (time)","level":2,"score":0.2572999894618988},{"id":"https://openalex.org/C105795698","wikidata":"https://www.wikidata.org/wiki/Q12483","display_name":"Statistics","level":1,"score":0.25380000472068787},{"id":"https://openalex.org/C87007009","wikidata":"https://www.wikidata.org/wiki/Q210832","display_name":"Statistical hypothesis testing","level":2,"score":0.25360000133514404}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2603.10848","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.10848","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"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.2603.10848","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.10848","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"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":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"In":[0,60],"Reinforcement":[1],"Learning":[2],"with":[3,80,97,172],"Verifiable":[4],"Rewards":[5],"(RLVR),":[6],"constructing":[7,134],"a":[8,78,90,104,135,173],"robust":[9,91],"advantage":[10],"baseline":[11,70,92,158],"is":[12],"critical":[13],"for":[14],"policy":[15,20,58,166],"gradients,":[16,167],"effectively":[17],"guiding":[18],"the":[19,48,53,57,69,81,114,122,129,140,145,157],"model":[21,44,55,75],"to":[22,50,138],"reinforce":[23],"desired":[24],"behaviors.":[25],"Recent":[26],"research":[27],"has":[28],"introduced":[29],"Generalist":[30],"Value":[31],"Models":[32],"(such":[33],"as":[34,77],"$V_0$),":[35],"which":[36,66],"achieve":[37],"pre-trained":[38],"value":[39,54,74,130],"estimation":[40],"by":[41,72,118],"explicitly":[42],"encoding":[43],"capabilities":[45],"in-context,":[46],"eliminating":[47],"need":[49],"synchronously":[51],"update":[52],"alongside":[56],"model.":[59],"this":[61],"paper,":[62],"we":[63,102],"propose":[64],"$V_{0.5}$,":[65],"adaptively":[67],"fuses":[68],"predicted":[71],"such":[73],"(acting":[76],"prior)":[79],"empirical":[82],"mean":[83],"derived":[84],"from":[85],"sparse":[86,119],"rollouts.":[87],"This":[88,112,154],"constructs":[89],"that":[93,186],"balances":[94,113],"computational":[95],"efficiency":[96],"extremely":[98],"low":[99],"variance.":[100],"Specifically,":[101],"introduce":[103],"real-time":[105],"statistical":[106],"testing":[107],"and":[108,191,196],"dynamic":[109],"budget":[110,151],"allocation.":[111],"high":[115],"variance":[116],"caused":[117],"sampling":[120],"against":[121],"systematic":[123],"bias":[124],"(or":[125],"hallucinations)":[126],"inherent":[127],"in":[128,143],"model's":[131],"prior.":[132],"By":[133],"hypothesis":[136],"test":[137],"evaluate":[139],"prior's":[141],"reliability":[142],"real-time,":[144],"system":[146],"dynamically":[147],"allocates":[148],"additional":[149],"rollout":[150],"on":[152],"demand.":[153],"mechanism":[155],"minimizes":[156],"estimator's":[159],"Mean":[160],"Squared":[161],"Error":[162],"(MSE),":[163],"guaranteeing":[164],"stable":[165],"even":[168],"under":[169],"extreme":[170],"sparsity":[171],"group":[174],"size":[175],"of":[176],"4.":[177],"Extensive":[178],"evaluations":[179],"across":[180],"six":[181],"mathematical":[182],"reasoning":[183],"benchmarks":[184],"demonstrate":[185],"$V_{0.5}$":[187],"significantly":[188],"outperforms":[189],"GRPO":[190],"DAPO,":[192],"achieving":[193],"faster":[194],"convergence":[195],"over":[197],"some":[198],"10%":[199],"performance":[200],"improvement.":[201]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-03-13T00:00:00"}