{"id":"https://openalex.org/W7167276854","doi":"https://doi.org/10.48550/arxiv.2607.01612","title":"Scaling with Confidence: Calibrating Confidence of LLMs for Adaptive Test Time Scaling","display_name":"Scaling with Confidence: Calibrating Confidence of LLMs for Adaptive Test Time Scaling","publication_year":2026,"publication_date":"2026-07-02","ids":{"openalex":"https://openalex.org/W7167276854","doi":"https://doi.org/10.48550/arxiv.2607.01612"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2607.01612","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2607.01612","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.2607.01612","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5139965117","display_name":"Xuqing Yang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yang, Xuqing","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5140010050","display_name":"Yi Yuan","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yuan, Yi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133170702","display_name":"Shanzhe Lei","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lei, Shanzhe","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5139979906","display_name":"Xuhong Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Xuhong","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":0.8303999900817871,"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":0.8303999900817871,"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":0.0364999994635582,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.03460000082850456,"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/calibration","display_name":"Calibration","score":0.698199987411499},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.6248999834060669},{"id":"https://openalex.org/keywords/test","display_name":"Test (biology)","score":0.5121999979019165},{"id":"https://openalex.org/keywords/scaling","display_name":"Scaling","score":0.4731000065803528},{"id":"https://openalex.org/keywords/reinforcement-learning","display_name":"Reinforcement learning","score":0.4708000123500824},{"id":"https://openalex.org/keywords/benchmark","display_name":"Benchmark (surveying)","score":0.36419999599456787},{"id":"https://openalex.org/keywords/confidence-interval","display_name":"Confidence interval","score":0.33739998936653137}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7287999987602234},{"id":"https://openalex.org/C165838908","wikidata":"https://www.wikidata.org/wiki/Q736777","display_name":"Calibration","level":2,"score":0.698199987411499},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.6248999834060669},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.6021000146865845},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.558899998664856},{"id":"https://openalex.org/C2777267654","wikidata":"https://www.wikidata.org/wiki/Q3519023","display_name":"Test (biology)","level":2,"score":0.5121999979019165},{"id":"https://openalex.org/C99844830","wikidata":"https://www.wikidata.org/wiki/Q102441924","display_name":"Scaling","level":2,"score":0.4731000065803528},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.4708000123500824},{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.36419999599456787},{"id":"https://openalex.org/C44249647","wikidata":"https://www.wikidata.org/wiki/Q208498","display_name":"Confidence interval","level":2,"score":0.33739998936653137},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.3025999963283539},{"id":"https://openalex.org/C520049643","wikidata":"https://www.wikidata.org/wiki/Q189760","display_name":"Voting","level":3,"score":0.29760000109672546},{"id":"https://openalex.org/C81917197","wikidata":"https://www.wikidata.org/wiki/Q628760","display_name":"Selection (genetic algorithm)","level":2,"score":0.28209999203681946},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.2709999978542328},{"id":"https://openalex.org/C16910744","wikidata":"https://www.wikidata.org/wiki/Q7705759","display_name":"Test data","level":2,"score":0.2703999876976013},{"id":"https://openalex.org/C2777868144","wikidata":"https://www.wikidata.org/wiki/Q7239817","display_name":"Preference elicitation","level":3,"score":0.27000001072883606},{"id":"https://openalex.org/C2909755999","wikidata":"https://www.wikidata.org/wiki/Q4751126","display_name":"Low Confidence","level":2,"score":0.25839999318122864},{"id":"https://openalex.org/C2911011789","wikidata":"https://www.wikidata.org/wiki/Q130741","display_name":"Hallucinating","level":2,"score":0.2565999925136566},{"id":"https://openalex.org/C12725497","wikidata":"https://www.wikidata.org/wiki/Q810247","display_name":"Baseline (sea)","level":2,"score":0.2540000081062317}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2607.01612","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2607.01612","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.2607.01612","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2607.01612","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":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Training":[0],"large":[1],"language":[2],"models":[3,31,56,188],"(LLMs)":[4],"with":[5],"reinforcement":[6],"learning":[7],"(RL)":[8],"has":[9],"significantly":[10],"advanced":[11],"their":[12,34],"performance":[13,43,111],"on":[14,140,150],"reasoning":[15],"and":[16,53,69,82,93,112,153,172,181,187],"question-answering":[17],"tasks.":[18],"However,":[19],"prevailing":[20],"RL":[21,77],"reward":[22],"designs":[23],"typically":[24],"prioritize":[25],"response":[26,141],"correctness,":[27,80],"neglecting":[28],"to":[29,32,39,57,163],"incentivize":[30],"express":[33],"confidence":[35,52,119],"accurately.":[36],"This":[37],"leads":[38],"a":[40,75],"critical":[41],"problem:":[42],"gains":[44],"are":[45],"often":[46],"accompanied":[47],"by":[48,161],"poor":[49],"calibration":[50,81,100,113],"between":[51],"accuracy,":[54,103],"misleading":[55],"overconfidently":[58],"hallucinate":[59],"when":[60],"uncertain.":[61],"To":[62],"address":[63],"this":[64],"limitation,":[65],"we":[66,122],"propose":[67],"$\\textbf{C}$orrectness":[68],"$\\textbf{C}$onfidence":[70],"$\\textbf{C}$alibration":[71],"$\\textbf{R}$einforcement":[72],"$\\textbf{L}$earning":[73],"($\\textbf{C3RL}$),":[74],"novel":[76],"algorithm":[78],"integrating":[79],"dataset-informed":[83],"reference":[84],"accuracy":[85],"rewards":[86],"together.":[87],"Comprehensive":[88],"evaluation":[89],"across":[90],"8":[91],"text":[92],"multimodal":[94],"datasets":[95,155],"demonstrates":[96],"that":[97,135,145],"C3RL":[98,171],"enhances":[99],"without":[101],"sacrificing":[102],"outperforming":[104],"the":[105,116,158,168,175],"current":[106],"state-of-the-art":[107],"method":[108],"in":[109],"both":[110,151],"metrics.":[114],"Utilizing":[115],"well-calibrated":[117],"verbalized":[118],"from":[120],"C3RL,":[121],"further":[123],"introduce":[124],"$\\textbf{C}$onfidence-based":[125],"$\\textbf{A}$daptive":[126],"Test":[127],"Time":[128],"$\\textbf{S}$caling":[129],"($\\textbf{CAS}$),":[130],"an":[131],"adjustable":[132],"inference-time":[133],"strategy":[134],"allocates":[136],"computational":[137],"resources":[138],"based":[139],"confidence.":[142],"Experiments":[143],"show":[144],"CAS":[146,173],"surpasses":[147],"majority":[148],"voting":[149],"in-domain":[152],"out-of-domain":[154],"while":[156],"reducing":[157],"inference":[159],"budget":[160],"up":[162],"12.33":[164],"times.":[165],"We":[166],"believe":[167],"synergy":[169],"of":[170],"paves":[174],"way":[176],"for":[177],"deploying":[178],"more":[179],"reliable":[180],"resource-efficient":[182],"LLMs.":[183],"The":[184],"code,":[185],"data":[186],"will":[189],"be":[190],"released.":[191]},"counts_by_year":[],"updated_date":"2026-07-28T07:46:37.118299","created_date":"2026-07-04T00:00:00"}