{"id":"https://openalex.org/W7128778850","doi":"https://doi.org/10.48550/arxiv.2602.11549","title":"Native Reasoning Models: Training Language Models to Reason on Unverifiable Data","display_name":"Native Reasoning Models: Training Language Models to Reason on Unverifiable Data","publication_year":2026,"publication_date":"2026-02-12","ids":{"openalex":"https://openalex.org/W7128778850","doi":"https://doi.org/10.48550/arxiv.2602.11549"},"language":null,"primary_location":{"id":"pmh:doi:10.48550/arxiv.2602.11549","is_oa":true,"landing_page_url":null,"pdf_url":null,"source":{"id":"https://openalex.org/S4406922384","display_name":"Open MIND","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Article"},"type":"article","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":null,"any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5111099262","display_name":"Yuanfu Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Yuanfu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5125934586","display_name":"Zhixuan Liu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liu, Zhixuan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5014052902","display_name":"Xiaoran Li","orcid":"https://orcid.org/0000-0002-5749-7615"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Xiangtian","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5125902154","display_name":"Chaochao Lu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lu, Chaochao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5125918915","display_name":"Chao Yang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yang, Chao","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":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.18689766,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"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":0.5383999943733215,"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":0.5383999943733215,"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":0.2540000081062317,"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/T12026","display_name":"Explainable Artificial Intelligence (XAI)","score":0.03750000149011612,"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/reinforcement-learning","display_name":"Reinforcement learning","score":0.6049000024795532},{"id":"https://openalex.org/keywords/scalability","display_name":"Scalability","score":0.5169000029563904},{"id":"https://openalex.org/keywords/perspective","display_name":"Perspective (graphical)","score":0.48739999532699585},{"id":"https://openalex.org/keywords/robustness","display_name":"Robustness (evolution)","score":0.4799000024795532},{"id":"https://openalex.org/keywords/process","display_name":"Process (computing)","score":0.46639999747276306},{"id":"https://openalex.org/keywords/commonsense-reasoning","display_name":"Commonsense reasoning","score":0.4433000087738037},{"id":"https://openalex.org/keywords/analytic-reasoning","display_name":"Analytic reasoning","score":0.4345000088214874},{"id":"https://openalex.org/keywords/embedding","display_name":"Embedding","score":0.3935999870300293},{"id":"https://openalex.org/keywords/language-model","display_name":"Language model","score":0.37529999017715454}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6973999738693237},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.6049000024795532},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.550599992275238},{"id":"https://openalex.org/C48044578","wikidata":"https://www.wikidata.org/wiki/Q727490","display_name":"Scalability","level":2,"score":0.5169000029563904},{"id":"https://openalex.org/C12713177","wikidata":"https://www.wikidata.org/wiki/Q1900281","display_name":"Perspective (graphical)","level":2,"score":0.48739999532699585},{"id":"https://openalex.org/C63479239","wikidata":"https://www.wikidata.org/wiki/Q7353546","display_name":"Robustness (evolution)","level":3,"score":0.4799000024795532},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.46639999747276306},{"id":"https://openalex.org/C193221554","wikidata":"https://www.wikidata.org/wiki/Q5153664","display_name":"Commonsense reasoning","level":2,"score":0.4433000087738037},{"id":"https://openalex.org/C103057564","wikidata":"https://www.wikidata.org/wiki/Q4751139","display_name":"Analytic reasoning","level":3,"score":0.4345000088214874},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.40849998593330383},{"id":"https://openalex.org/C41608201","wikidata":"https://www.wikidata.org/wiki/Q980509","display_name":"Embedding","level":2,"score":0.3935999870300293},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.37529999017715454},{"id":"https://openalex.org/C77618280","wikidata":"https://www.wikidata.org/wiki/Q1155772","display_name":"Scheme (mathematics)","level":2,"score":0.3612000048160553},{"id":"https://openalex.org/C89288958","wikidata":"https://www.wikidata.org/wiki/Q7301504","display_name":"Reasoning system","level":2,"score":0.3357999920845032},{"id":"https://openalex.org/C166088908","wikidata":"https://www.wikidata.org/wiki/Q308495","display_name":"Abductive reasoning","level":2,"score":0.319599986076355},{"id":"https://openalex.org/C83725634","wikidata":"https://www.wikidata.org/wiki/Q7268699","display_name":"Qualitative reasoning","level":2,"score":0.2996000051498413},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.296999990940094},{"id":"https://openalex.org/C204323151","wikidata":"https://www.wikidata.org/wiki/Q905424","display_name":"Range (aeronautics)","level":2,"score":0.2921000123023987},{"id":"https://openalex.org/C169900460","wikidata":"https://www.wikidata.org/wiki/Q2200417","display_name":"Cognition","level":2,"score":0.28679999709129333},{"id":"https://openalex.org/C2781067378","wikidata":"https://www.wikidata.org/wiki/Q17027399","display_name":"Interpretability","level":2,"score":0.2800000011920929},{"id":"https://openalex.org/C85847156","wikidata":"https://www.wikidata.org/wiki/Q59015987","display_name":"Verifiable secret sharing","level":3,"score":0.2750999927520752},{"id":"https://openalex.org/C195344581","wikidata":"https://www.wikidata.org/wiki/Q2555318","display_name":"Automated reasoning","level":2,"score":0.2750999927520752},{"id":"https://openalex.org/C115086926","wikidata":"https://www.wikidata.org/wiki/Q17004651","display_name":"Causal reasoning","level":3,"score":0.27379998564720154},{"id":"https://openalex.org/C2777508537","wikidata":"https://www.wikidata.org/wiki/Q7936620","display_name":"Visual reasoning","level":2,"score":0.26989999413490295},{"id":"https://openalex.org/C20162079","wikidata":"https://www.wikidata.org/wiki/Q1151406","display_name":"Case-based reasoning","level":2,"score":0.26269999146461487}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:doi:10.48550/arxiv.2602.11549","is_oa":true,"landing_page_url":null,"pdf_url":null,"source":{"id":"https://openalex.org/S4406922384","display_name":"Open MIND","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Article"},{"id":"doi:10.48550/arxiv.2602.11549","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2602.11549","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":"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":"pmh:doi:10.48550/arxiv.2602.11549","is_oa":true,"landing_page_url":null,"pdf_url":null,"source":{"id":"https://openalex.org/S4406922384","display_name":"Open MIND","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Article"},"sustainable_development_goals":[{"score":0.431015282869339,"id":"https://metadata.un.org/sdg/4","display_name":"Quality Education"}],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"The":[0],"prevailing":[1],"paradigm":[2],"for":[3,101],"training":[4,107,122],"large":[5],"reasoning":[6,26,82,90,112,126,223,244],"models--combining":[7],"Supervised":[8],"Fine-Tuning":[9],"(SFT)":[10],"with":[11,14],"Reinforcement":[12],"Learning":[13],"Verifiable":[15],"Rewards":[16],"(RLVR)--is":[17],"fundamentally":[18],"constrained":[19],"by":[20,83,109],"its":[21,64,88,184],"reliance":[22],"on":[23,189],"high-quality,":[24],"human-annotated":[25],"data":[27],"and":[28,42,54,161,191,209,225,241],"external":[29],"verifiers.":[30],"This":[31,144],"dependency":[32],"incurs":[33],"significant":[34],"data-collection":[35],"costs,":[36],"risks":[37],"embedding":[38],"human":[39],"cognitive":[40],"biases,":[41],"confines":[43],"the":[44,85,99,106,111,136,141,175],"reinforcement":[45],"learning":[46],"stage":[47],"to":[48,149,178,229],"objectively":[49],"assessable":[50],"domains":[51,224],"like":[52],"mathematics":[53],"coding,":[55],"leaving":[56],"a":[57,76,115,120,170,233],"wide":[58],"range":[59],"of":[60,139,154],"unverifiable":[61],"tasks":[62],"beyond":[63],"scope.":[65],"To":[66],"overcome":[67],"these":[68],"limitations,":[69],"we":[70],"introduce":[71],"NRT":[72,104,197],"(Native":[73],"Reasoning":[74],"Training),":[75],"novel":[77],"framework":[78],"that":[79,124,134,182,196],"cultivates":[80],"complex":[81,222],"having":[84],"model":[86,176,193],"generate":[87],"own":[89,185],"traces":[91],"using":[92],"only":[93],"standard":[94,206],"question-answer":[95],"pairs,":[96],"thereby":[97],"obviating":[98],"need":[100],"expert-written":[102],"demonstrations.":[103],"reframes":[105],"problem":[108],"treating":[110],"process":[113],"as":[114,127,158],"latent":[116],"variable.":[117],"It":[118],"employs":[119],"unified":[121,145],"objective":[123],"models":[125],"an":[128],"optimization":[129],"problem,":[130],"intrinsically":[131],"rewarding":[132],"paths":[133],"increase":[135],"model's":[137],"likelihood":[138],"producing":[140],"ground-truth":[142],"answer.":[143],"perspective":[146],"allows":[147],"us":[148],"analyze":[150],"intrinsic":[151],"failure":[152],"modes":[153],"prior":[155,210],"methods,":[156,203],"such":[157],"policy":[159,230],"collapse,":[160,231],"systematically":[162],"design":[163],"more":[164,239],"robust":[165],"reward":[166],"aggregation":[167],"functions,":[168],"creating":[169],"self-reinforcing":[171],"feedback":[172],"loop":[173],"where":[174],"learns":[177],"think":[179],"in":[180,221],"ways":[181],"resolve":[183],"uncertainty.":[186],"Empirical":[187],"evaluation":[188],"Llama":[190],"Mistral":[192],"families":[194],"demonstrates":[195],"achieves":[198],"state-of-the-art":[199],"performance":[200,219],"among":[201],"verifier-free":[202,211],"significantly":[204],"outperforming":[205],"SFT":[207],"baselines":[208],"RL":[212],"methods.":[213],"Our":[214],"approach":[215],"yields":[216],"particularly":[217],"strong":[218],"gains":[220],"exhibits":[226],"high":[227],"robustness":[228],"offering":[232],"general,":[234],"scalable":[235],"path":[236],"toward":[237],"building":[238],"powerful":[240],"broadly":[242],"applicable":[243],"systems.":[245]},"counts_by_year":[],"updated_date":"2026-07-15T18:14:33.161393","created_date":"2026-02-14T00:00:00"}