{"id":"https://openalex.org/W7140198094","doi":"https://doi.org/10.48550/arxiv.2603.20314","title":"VGS-Decoding: Visual Grounding Score Guided Decoding for Hallucination Mitigation in Medical VLMs","display_name":"VGS-Decoding: Visual Grounding Score Guided Decoding for Hallucination Mitigation in Medical VLMs","publication_year":2026,"publication_date":"2026-03-19","ids":{"openalex":"https://openalex.org/W7140198094","doi":"https://doi.org/10.48550/arxiv.2603.20314"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2603.20314","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.20314","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.2603.20314","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":null,"display_name":"Kolli, Govinda","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Kolli, Govinda","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Dukre, Adinath Madhavrao","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Dukre, Adinath Madhavrao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Bozorgtabar, Behzad","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Bozorgtabar, Behzad","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Mahapatra, Dwarikanath","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Mahapatra, Dwarikanath","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":null,"display_name":"Razzak, Imran","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Razzak, Imran","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/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.3682999908924103,"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/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.3682999908924103,"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":0.15639999508857727,"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/T11775","display_name":"COVID-19 diagnosis using AI","score":0.04230000078678131,"subfield":{"id":"https://openalex.org/subfields/2741","display_name":"Radiology, Nuclear Medicine and Imaging"},"field":{"id":"https://openalex.org/fields/27","display_name":"Medicine"},"domain":{"id":"https://openalex.org/domains/4","display_name":"Health Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/hallucinating","display_name":"Hallucinating","score":0.9333999752998352},{"id":"https://openalex.org/keywords/decoding-methods","display_name":"Decoding methods","score":0.613099992275238},{"id":"https://openalex.org/keywords/visual-hallucination","display_name":"Visual Hallucination","score":0.5546000003814697},{"id":"https://openalex.org/keywords/artifact","display_name":"Artifact (error)","score":0.5412999987602234},{"id":"https://openalex.org/keywords/prior-probability","display_name":"Prior probability","score":0.5306000113487244},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.5218999981880188},{"id":"https://openalex.org/keywords/code","display_name":"Code (set theory)","score":0.4359000027179718},{"id":"https://openalex.org/keywords/dependency","display_name":"Dependency (UML)","score":0.41679999232292175},{"id":"https://openalex.org/keywords/overhead","display_name":"Overhead (engineering)","score":0.3955000042915344}],"concepts":[{"id":"https://openalex.org/C2911011789","wikidata":"https://www.wikidata.org/wiki/Q130741","display_name":"Hallucinating","level":2,"score":0.9333999752998352},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6762999892234802},{"id":"https://openalex.org/C57273362","wikidata":"https://www.wikidata.org/wiki/Q576722","display_name":"Decoding methods","level":2,"score":0.613099992275238},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5971999764442444},{"id":"https://openalex.org/C2908998935","wikidata":"https://www.wikidata.org/wiki/Q130741","display_name":"Visual Hallucination","level":2,"score":0.5546000003814697},{"id":"https://openalex.org/C2779010991","wikidata":"https://www.wikidata.org/wiki/Q2720909","display_name":"Artifact (error)","level":2,"score":0.5412999987602234},{"id":"https://openalex.org/C177769412","wikidata":"https://www.wikidata.org/wiki/Q278090","display_name":"Prior probability","level":3,"score":0.5306000113487244},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.5218999981880188},{"id":"https://openalex.org/C2776760102","wikidata":"https://www.wikidata.org/wiki/Q5139990","display_name":"Code (set theory)","level":3,"score":0.4359000027179718},{"id":"https://openalex.org/C19768560","wikidata":"https://www.wikidata.org/wiki/Q320727","display_name":"Dependency (UML)","level":2,"score":0.41679999232292175},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.41029998660087585},{"id":"https://openalex.org/C2779960059","wikidata":"https://www.wikidata.org/wiki/Q7113681","display_name":"Overhead (engineering)","level":2,"score":0.3955000042915344},{"id":"https://openalex.org/C178253425","wikidata":"https://www.wikidata.org/wiki/Q162668","display_name":"Visual perception","level":3,"score":0.3894999921321869},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.38839998841285706},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.3806999921798706},{"id":"https://openalex.org/C168993435","wikidata":"https://www.wikidata.org/wiki/Q6501125","display_name":"Ground","level":2,"score":0.36640000343322754},{"id":"https://openalex.org/C125411270","wikidata":"https://www.wikidata.org/wiki/Q18653","display_name":"Encoding (memory)","level":2,"score":0.35899999737739563},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.3564999997615814},{"id":"https://openalex.org/C36464697","wikidata":"https://www.wikidata.org/wiki/Q451553","display_name":"Visualization","level":2,"score":0.3434000015258789},{"id":"https://openalex.org/C160234255","wikidata":"https://www.wikidata.org/wiki/Q812535","display_name":"Bayesian inference","level":3,"score":0.34209999442100525},{"id":"https://openalex.org/C31972630","wikidata":"https://www.wikidata.org/wiki/Q844240","display_name":"Computer vision","level":1,"score":0.32829999923706055},{"id":"https://openalex.org/C114289077","wikidata":"https://www.wikidata.org/wiki/Q3284399","display_name":"Statistical model","level":2,"score":0.3264999985694885},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.3215000033378601},{"id":"https://openalex.org/C2777508537","wikidata":"https://www.wikidata.org/wiki/Q7936620","display_name":"Visual reasoning","level":2,"score":0.2791000008583069},{"id":"https://openalex.org/C134261354","wikidata":"https://www.wikidata.org/wiki/Q938438","display_name":"Statistical inference","level":2,"score":0.2702000141143799},{"id":"https://openalex.org/C2777402240","wikidata":"https://www.wikidata.org/wiki/Q6783436","display_name":"Masking (illustration)","level":2,"score":0.26660001277923584},{"id":"https://openalex.org/C15744967","wikidata":"https://www.wikidata.org/wiki/Q9418","display_name":"Psychology","level":0,"score":0.2565000057220459},{"id":"https://openalex.org/C49937458","wikidata":"https://www.wikidata.org/wiki/Q2599292","display_name":"Probabilistic logic","level":2,"score":0.25450000166893005},{"id":"https://openalex.org/C160086991","wikidata":"https://www.wikidata.org/wiki/Q5939193","display_name":"Human visual system model","level":3,"score":0.2524999976158142},{"id":"https://openalex.org/C26760741","wikidata":"https://www.wikidata.org/wiki/Q160402","display_name":"Perception","level":2,"score":0.250900000333786},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.25049999356269836}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2603.20314","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.20314","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.2603.20314","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.20314","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":{"Medical":[0],"Vision-Language":[1],"Models":[2],"(VLMs)":[3],"often":[4],"hallucinate":[5],"by":[6,75,88],"generating":[7],"responses":[8],"based":[9],"on":[10,106],"language":[11],"priors":[12],"rather":[13],"than":[14],"visual":[15,51,73],"evidence,":[16],"posing":[17],"risks":[18],"in":[19,60,126],"clinical":[20,143],"applications.":[21],"We":[22,62],"propose":[23],"Visual":[24,65],"Grounding":[25,66],"Score":[26,67],"Guided":[27],"Decoding":[28],"(VGS-Decoding),":[29],"a":[30],"training-free":[31],"method":[32],"to":[33,120,152],"mitigate":[34],"hallucinations":[35],"during":[36],"inference.":[37],"Our":[38],"key":[39],"insight":[40],"is":[41,53],"that":[42],"hallucinated":[43],"tokens":[44,58,92],"maintain":[45],"or":[46],"increase":[47],"their":[48],"probability":[49],"when":[50],"information":[52],"degraded,":[54],"while":[55,93,129],"visually":[56,90],"grounded":[57,91],"decrease":[59],"probability.":[61],"introduce":[63],"the":[64],"(VGS),":[68],"which":[69],"measures":[70],"each":[71],"token's":[72],"dependency":[74],"comparing":[76],"distributions":[77],"from":[78],"original":[79],"and":[80,108,113,124,135],"distorted":[81],"images.":[82],"During":[83],"decoding,":[84],"we":[85],"reweight":[86],"probabilities":[87],"amplifying":[89],"suppressing":[94],"hallucinations.":[95],"Unlike":[96],"fixed-weight":[97],"contrastive":[98],"methods,":[99],"VGS-Decoding":[100],"provides":[101],"per-token":[102],"adaptive":[103],"control.":[104],"Experiments":[105],"MIMIC-Diff-VQA":[107],"VQA-RAD":[109],"across":[110],"LLaVA-Med,":[111],"CheXagent,":[112],"MedGemma":[114],"demonstrate":[115],"consistent":[116],"improvements,":[117],"with":[118],"up":[119],"+9.12%":[121],"overall":[122],"gain":[123],"$+8.98\\%$":[125],"open-ended":[127],"recall,":[128],"introducing":[130],"only":[131],"$2\\times$":[132],"inference":[133],"overhead":[134],"no":[136],"additional":[137],"training,":[138],"making":[139],"it":[140],"practical":[141],"for":[142],"deployment.":[144],"Upon":[145],"acceptance,":[146],"code":[147],"will":[148],"be":[149],"released":[150],"publicly":[151],"facilitate":[153],"reproducibility.":[154]},"counts_by_year":[],"updated_date":"2026-07-28T07:46:37.118299","created_date":"2026-03-25T00:00:00"}