{"id":"https://openalex.org/W7151094253","doi":"https://doi.org/10.48550/arxiv.2604.02543","title":"Overconfidence and Calibration in Medical VQA: Empirical Findings and Hallucination-Aware Mitigation","display_name":"Overconfidence and Calibration in Medical VQA: Empirical Findings and Hallucination-Aware Mitigation","publication_year":2026,"publication_date":"2026-04-02","ids":{"openalex":"https://openalex.org/W7151094253","doi":"https://doi.org/10.48550/arxiv.2604.02543"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2604.02543","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.02543","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":"article"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2604.02543","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5063529811","display_name":"Joongmoo Byun","orcid":"https://orcid.org/0000-0003-0445-0271"},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Byun, Ji Young","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133031478","display_name":"Young-Jin Park","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Park, Young-Jin","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5059097691","display_name":"Jean-Philippe Corbeil","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Corbeil, Jean-Philippe","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5133021950","display_name":"Asma Ben Abacha","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Abacha, Asma Ben","raw_affiliation_strings":[],"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":4,"corresponding_author_ids":["https://openalex.org/A5063529811"],"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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.9269999861717224,"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"}},"topics":[{"id":"https://openalex.org/T11714","display_name":"Multimodal Machine Learning Applications","score":0.9269999861717224,"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":0.01600000075995922,"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":0.01549999974668026,"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/overconfidence-effect","display_name":"Overconfidence effect","score":0.9200000166893005},{"id":"https://openalex.org/keywords/calibration","display_name":"Calibration","score":0.791100025177002},{"id":"https://openalex.org/keywords/discriminative-model","display_name":"Discriminative model","score":0.7145000100135803},{"id":"https://openalex.org/keywords/confidence-interval","display_name":"Confidence interval","score":0.573199987411499},{"id":"https://openalex.org/keywords/baseline","display_name":"Baseline (sea)","score":0.4611999988555908},{"id":"https://openalex.org/keywords/medical-diagnosis","display_name":"Medical diagnosis","score":0.4083000123500824},{"id":"https://openalex.org/keywords/quality","display_name":"Quality (philosophy)","score":0.396699994802475},{"id":"https://openalex.org/keywords/empirical-research","display_name":"Empirical research","score":0.38839998841285706},{"id":"https://openalex.org/keywords/key","display_name":"Key (lock)","score":0.35760000348091125}],"concepts":[{"id":"https://openalex.org/C51110983","wikidata":"https://www.wikidata.org/wiki/Q16503490","display_name":"Overconfidence effect","level":2,"score":0.9200000166893005},{"id":"https://openalex.org/C165838908","wikidata":"https://www.wikidata.org/wiki/Q736777","display_name":"Calibration","level":2,"score":0.791100025177002},{"id":"https://openalex.org/C97931131","wikidata":"https://www.wikidata.org/wiki/Q5282087","display_name":"Discriminative model","level":2,"score":0.7145000100135803},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.613099992275238},{"id":"https://openalex.org/C44249647","wikidata":"https://www.wikidata.org/wiki/Q208498","display_name":"Confidence interval","level":2,"score":0.573199987411499},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.4945000112056732},{"id":"https://openalex.org/C149782125","wikidata":"https://www.wikidata.org/wiki/Q160039","display_name":"Econometrics","level":1,"score":0.4934999942779541},{"id":"https://openalex.org/C12725497","wikidata":"https://www.wikidata.org/wiki/Q810247","display_name":"Baseline (sea)","level":2,"score":0.4611999988555908},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4593999981880188},{"id":"https://openalex.org/C105795698","wikidata":"https://www.wikidata.org/wiki/Q12483","display_name":"Statistics","level":1,"score":0.43880000710487366},{"id":"https://openalex.org/C534262118","wikidata":"https://www.wikidata.org/wiki/Q177719","display_name":"Medical diagnosis","level":2,"score":0.4083000123500824},{"id":"https://openalex.org/C2779530757","wikidata":"https://www.wikidata.org/wiki/Q1207505","display_name":"Quality (philosophy)","level":2,"score":0.396699994802475},{"id":"https://openalex.org/C120936955","wikidata":"https://www.wikidata.org/wiki/Q2155640","display_name":"Empirical research","level":2,"score":0.38839998841285706},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.36079999804496765},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.35760000348091125},{"id":"https://openalex.org/C2779351106","wikidata":"https://www.wikidata.org/wiki/Q24965456","display_name":"Decision quality","level":3,"score":0.32580000162124634},{"id":"https://openalex.org/C75291252","wikidata":"https://www.wikidata.org/wiki/Q1315756","display_name":"TRACE (psycholinguistics)","level":2,"score":0.31700000166893005},{"id":"https://openalex.org/C2983241795","wikidata":"https://www.wikidata.org/wiki/Q6806500","display_name":"Medical decision making","level":2,"score":0.30070000886917114},{"id":"https://openalex.org/C15744967","wikidata":"https://www.wikidata.org/wiki/Q9418","display_name":"Psychology","level":0,"score":0.29580000042915344},{"id":"https://openalex.org/C166052673","wikidata":"https://www.wikidata.org/wiki/Q83021","display_name":"Empirical evidence","level":2,"score":0.2919999957084656},{"id":"https://openalex.org/C96250715","wikidata":"https://www.wikidata.org/wiki/Q965330","display_name":"Estimation","level":2,"score":0.29109999537467957},{"id":"https://openalex.org/C81917197","wikidata":"https://www.wikidata.org/wiki/Q628760","display_name":"Selection (genetic algorithm)","level":2,"score":0.2867000102996826},{"id":"https://openalex.org/C66520545","wikidata":"https://www.wikidata.org/wiki/Q7353538","display_name":"Robust confidence intervals","level":3,"score":0.2775000035762787},{"id":"https://openalex.org/C2778755073","wikidata":"https://www.wikidata.org/wiki/Q10858537","display_name":"Scale (ratio)","level":2,"score":0.2712000012397766},{"id":"https://openalex.org/C13662513","wikidata":"https://www.wikidata.org/wiki/Q5160087","display_name":"Confidence distribution","level":3,"score":0.2662999927997589},{"id":"https://openalex.org/C105339364","wikidata":"https://www.wikidata.org/wiki/Q2297740","display_name":"Software deployment","level":2,"score":0.26109999418258667},{"id":"https://openalex.org/C151201525","wikidata":"https://www.wikidata.org/wiki/Q177239","display_name":"Limit (mathematics)","level":2,"score":0.258899986743927},{"id":"https://openalex.org/C118147538","wikidata":"https://www.wikidata.org/wiki/Q1126970","display_name":"Metacognition","level":3,"score":0.2549000084400177}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2604.02543","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.02543","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":"article"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2604.02543","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.02543","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":"article"},"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/10","score":0.6572712659835815,"display_name":"Reduced inequalities"}],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"As":[0],"vision-language":[1],"models":[2,36],"(VLMs)":[3],"are":[4,138],"increasingly":[5],"deployed":[6],"in":[7,40,56,141,227],"clinical":[8],"decision":[9],"support,":[10],"more":[11,222],"than":[12],"accuracy":[13],"is":[14,22,96],"required:":[15],"knowing":[16],"when":[17],"to":[18,130,172,220],"trust":[19],"their":[20,131],"predictions":[21],"equally":[23],"critical.":[24],"Yet,":[25],"a":[26,49],"comprehensive":[27,50],"and":[28,69,95,106,122,186,212],"systematic":[29],"investigation":[30],"into":[31],"the":[32,41,125,143,151,189,214],"overconfidence":[33,90],"of":[34,53,146,217,225],"these":[35,134,156,180],"remains":[37],"notably":[38],"scarce":[39],"medical":[42,77,205,228],"domain.":[43],"We":[44,176],"address":[45],"this":[46],"gap":[47],"through":[48],"empirical":[51],"study":[52,84],"confidence":[54,71,108,174,210],"calibration":[55,113,120,136,161,185,200],"VLMs,":[57],"spanning":[58],"three":[59,65,76,86],"model":[60,66,93],"families":[61,94],"(Qwen3-VL,":[62],"InternVL3,":[63],"LLaVA-NeXT),":[64],"scales":[67],"(2B--38B),":[68],"multiple":[70],"estimation":[72],"prompting":[73],"strategies,":[74],"across":[75,92],"visual":[78],"question":[79],"answering":[80],"(VQA)":[81],"benchmarks.":[82],"Our":[83],"yields":[85],"key":[87],"findings:":[88],"First,":[89],"persists":[91],"not":[97],"resolved":[98],"by":[99,155],"scaling":[100],"or":[101],"prompting,":[102],"such":[103,115],"as":[104,116,169,201],"chain-of-thought":[105],"verbalized":[107],"variants.":[109],"Second,":[110],"simple":[111],"post-hoc":[112,135,199],"approaches,":[114],"Platt":[117],"scaling,":[118],"reduce":[119],"error":[121],"consistently":[123],"outperform":[124],"prompt-based":[126],"strategy.":[127],"Third,":[128],"due":[129],"(strict)":[132],"monotonicity,":[133],"methods":[137],"inherently":[139],"limited":[140],"improving":[142],"discriminative":[144],"quality":[145],"predictions,":[147],"leaving":[148],"AUROC":[149],"at":[150],"same":[152],"level.":[153],"Motivated":[154],"findings,":[157],"we":[158],"investigate":[159],"hallucination-aware":[160],"(HAC),":[162],"which":[163],"incorporates":[164],"vision-grounded":[165],"hallucination":[166,181,218],"detection":[167],"signals":[168,182,219],"complementary":[170],"inputs":[171],"refine":[173],"estimates.":[175],"find":[177],"that":[178],"leveraging":[179],"improves":[183],"both":[184],"AUROC,":[187],"with":[188],"largest":[190],"gains":[191],"on":[192],"open-ended":[193],"questions.":[194],"Overall,":[195],"our":[196],"findings":[197],"suggest":[198],"standard":[202],"practice":[203],"for":[204],"VLM":[206],"deployment":[207],"over":[208],"raw":[209],"estimates,":[211],"highlight":[213],"practical":[215],"usefulness":[216],"enable":[221],"reliable":[223],"use":[224],"VLMs":[226],"VQA.":[229]},"counts_by_year":[],"updated_date":"2026-04-07T06:06:30.997549","created_date":"2026-04-07T00:00:00"}