{"id":"https://openalex.org/W7156996919","doi":"https://doi.org/10.48550/arxiv.2604.22989","title":"CheXmix: Unified Generative Pretraining for Vision Language Models in Medical Imaging","display_name":"CheXmix: Unified Generative Pretraining for Vision Language Models in Medical Imaging","publication_year":2026,"publication_date":"2026-04-24","ids":{"openalex":"https://openalex.org/W7156996919","doi":"https://doi.org/10.48550/arxiv.2604.22989"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2604.22989","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.22989","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":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.2604.22989","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5134756757","display_name":"Ashwin Kumar","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Kumar, Ashwin","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5081687968","display_name":"Robbie Holland","orcid":"https://orcid.org/0000-0001-8373-8390"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Holland, Robbie","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134800668","display_name":"Corey Barrett","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Barrett, Corey","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5026474173","display_name":"Jangwon Kim","orcid":"https://orcid.org/0000-0003-0228-3502"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Kim, Jangwon","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5012771499","display_name":"Maya Varma","orcid":"https://orcid.org/0000-0003-0693-7753"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Varma, Maya","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134769108","display_name":"Zhihong Chen","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chen, Zhihong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134750833","display_name":"Yunhe Gao","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Gao, Yunhe","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5065577825","display_name":"Greg Zaharchuk","orcid":"https://orcid.org/0000-0001-5781-8848"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zaharchuk, Greg","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134762807","display_name":"Tara Taghavi","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Taghavi, Tara","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5002843568","display_name":"Krishnaram Kenthapadi","orcid":"https://orcid.org/0000-0003-1237-087X"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Kenthapadi, Krishnaram","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5134820046","display_name":"Akshay Chaudhari","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chaudhari, Akshay","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/T11775","display_name":"COVID-19 diagnosis using AI","score":0.4666999876499176,"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"}},"topics":[{"id":"https://openalex.org/T11775","display_name":"COVID-19 diagnosis using AI","score":0.4666999876499176,"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"}},{"id":"https://openalex.org/T11714","display_name":"Multimodal Machine Learning Applications","score":0.2386000007390976,"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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.050599999725818634,"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/discriminative-model","display_name":"Discriminative model","score":0.7070000171661377},{"id":"https://openalex.org/keywords/generative-model","display_name":"Generative model","score":0.6811000108718872},{"id":"https://openalex.org/keywords/generative-grammar","display_name":"Generative grammar","score":0.5809000134468079},{"id":"https://openalex.org/keywords/masking","display_name":"Masking (illustration)","score":0.5156000256538391},{"id":"https://openalex.org/keywords/projection","display_name":"Projection (relational algebra)","score":0.4745999872684479},{"id":"https://openalex.org/keywords/representation","display_name":"Representation (politics)","score":0.4230000078678131},{"id":"https://openalex.org/keywords/medical-imaging","display_name":"Medical imaging","score":0.40540000796318054},{"id":"https://openalex.org/keywords/salient","display_name":"Salient","score":0.40540000796318054},{"id":"https://openalex.org/keywords/encoder","display_name":"Encoder","score":0.3741999864578247}],"concepts":[{"id":"https://openalex.org/C97931131","wikidata":"https://www.wikidata.org/wiki/Q5282087","display_name":"Discriminative model","level":2,"score":0.7070000171661377},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7009999752044678},{"id":"https://openalex.org/C167966045","wikidata":"https://www.wikidata.org/wiki/Q5532625","display_name":"Generative model","level":3,"score":0.6811000108718872},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.6611999869346619},{"id":"https://openalex.org/C39890363","wikidata":"https://www.wikidata.org/wiki/Q36108","display_name":"Generative grammar","level":2,"score":0.5809000134468079},{"id":"https://openalex.org/C2777402240","wikidata":"https://www.wikidata.org/wiki/Q6783436","display_name":"Masking (illustration)","level":2,"score":0.5156000256538391},{"id":"https://openalex.org/C57493831","wikidata":"https://www.wikidata.org/wiki/Q3134666","display_name":"Projection (relational algebra)","level":2,"score":0.4745999872684479},{"id":"https://openalex.org/C2776359362","wikidata":"https://www.wikidata.org/wiki/Q2145286","display_name":"Representation (politics)","level":3,"score":0.4230000078678131},{"id":"https://openalex.org/C31601959","wikidata":"https://www.wikidata.org/wiki/Q931309","display_name":"Medical imaging","level":2,"score":0.40540000796318054},{"id":"https://openalex.org/C2780719617","wikidata":"https://www.wikidata.org/wiki/Q1030752","display_name":"Salient","level":2,"score":0.40540000796318054},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.38690000772476196},{"id":"https://openalex.org/C118505674","wikidata":"https://www.wikidata.org/wiki/Q42586063","display_name":"Encoder","level":2,"score":0.3741999864578247},{"id":"https://openalex.org/C176217482","wikidata":"https://www.wikidata.org/wiki/Q860554","display_name":"Metric (unit)","level":2,"score":0.36570000648498535},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.36500000953674316},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.3450999855995178},{"id":"https://openalex.org/C157657479","wikidata":"https://www.wikidata.org/wiki/Q2367247","display_name":"Closed captioning","level":3,"score":0.32919999957084656},{"id":"https://openalex.org/C59404180","wikidata":"https://www.wikidata.org/wiki/Q17013334","display_name":"Feature learning","level":2,"score":0.32739999890327454},{"id":"https://openalex.org/C64543145","wikidata":"https://www.wikidata.org/wiki/Q162942","display_name":"Intersection (aeronautics)","level":2,"score":0.3156000077724457},{"id":"https://openalex.org/C177769412","wikidata":"https://www.wikidata.org/wiki/Q278090","display_name":"Prior probability","level":3,"score":0.31130000948905945},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.30070000886917114},{"id":"https://openalex.org/C195324797","wikidata":"https://www.wikidata.org/wiki/Q33742","display_name":"Natural language","level":2,"score":0.2831000089645386},{"id":"https://openalex.org/C31972630","wikidata":"https://www.wikidata.org/wiki/Q844240","display_name":"Computer vision","level":1,"score":0.2815000116825104},{"id":"https://openalex.org/C70437156","wikidata":"https://www.wikidata.org/wiki/Q7228652","display_name":"Pooling","level":2,"score":0.2799000144004822},{"id":"https://openalex.org/C2776760102","wikidata":"https://www.wikidata.org/wiki/Q5139990","display_name":"Code (set theory)","level":3,"score":0.2775999903678894},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.2768999934196472},{"id":"https://openalex.org/C2780660688","wikidata":"https://www.wikidata.org/wiki/Q25052564","display_name":"Multimodal learning","level":2,"score":0.2718000113964081},{"id":"https://openalex.org/C115961682","wikidata":"https://www.wikidata.org/wiki/Q860623","display_name":"Image (mathematics)","level":2,"score":0.25619998574256897},{"id":"https://openalex.org/C2780513914","wikidata":"https://www.wikidata.org/wiki/Q18210350","display_name":"Bottleneck","level":2,"score":0.251800000667572}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2604.22989","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.22989","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":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.2604.22989","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.22989","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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"sustainable_development_goals":[{"display_name":"Reduced inequalities","score":0.7159584164619446,"id":"https://metadata.un.org/sdg/10"}],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Recent":[0],"medical":[1,41],"multimodal":[2,8,116],"foundation":[3],"models":[4,132,153,187],"are":[5,46,133],"built":[6],"as":[7,57],"LLMs":[9],"(MLLMs)":[10],"by":[11,63,112,158,163,191],"connecting":[12],"a":[13,28,70,89,96,114,210],"CLIP-pretrained":[14],"vision":[15],"encoder":[16],"to":[17],"an":[18],"LLM":[19],"using":[20],"LLaVA-style":[21],"finetuning.":[22],"This":[23,36],"two-stage,":[24],"decoupled":[25],"approach":[26,149],"introduces":[27],"projection":[29,61],"layer":[30],"that":[31,78,120,204],"can":[32],"distort":[33],"visual":[34],"features.":[35],"is":[37,219],"especially":[38],"concerning":[39],"in":[40],"imaging":[42],"where":[43],"subtle":[44],"cues":[45],"essential":[47],"for":[48,197],"accurate":[49],"diagnoses.":[50],"In":[51],"contrast,":[52],"early-fusion":[53,91],"generative":[54,92,117,140,152,186],"approaches":[55],"such":[56],"Chameleon":[58],"eliminate":[59],"the":[60,80,122,173,194],"bottleneck":[62],"processing":[64],"image":[65,169],"and":[66,139,145,160,188],"text":[67],"tokens":[68],"within":[69],"single":[71],"unified":[72,90],"sequence,":[73],"enabling":[74],"joint":[75],"representation":[76],"learning":[77],"leverages":[79],"inductive":[81],"priors":[82],"of":[83,99,125,213],"language":[84],"models.":[85],"We":[86,106,177],"present":[87],"CheXmix,":[88],"model":[93],"trained":[94],"on":[95,108,165,172,193],"large":[97],"corpus":[98],"chest":[100,214],"X-rays":[101],"paired":[102],"with":[103,128],"radiology":[104,198],"reports.":[105],"expand":[107],"Chameleon's":[109],"autoregressive":[110],"framework":[111],"introducing":[113],"two-stage":[115],"pretraining":[118],"strategy":[119],"combines":[121],"representational":[123],"strengths":[124],"masked":[126],"autoencoders":[127],"MLLMs.":[129],"The":[130],"resulting":[131],"highly":[134],"flexible,":[135],"supporting":[136],"both":[137,143],"discriminative":[138],"tasks":[141],"at":[142,167],"coarse":[144],"fine-grained":[146,207],"scales.":[147],"Our":[148,217],"outperforms":[150],"well-established":[151],"across":[154,209],"all":[155],"masking":[156,170],"ratios":[157,171],"6.0%":[159],"surpasses":[161],"CheXagent":[162,190],"8.6%":[164],"AUROC":[166],"high":[168],"CheXpert":[174],"classification":[175],"task.":[176],"further":[178],"inpaint":[179],"images":[180],"over":[181],"51.0%":[182],"better":[183],"than":[184],"text-only":[185],"outperform":[189],"45%":[192],"GREEN":[195],"metric":[196],"report":[199],"generation.":[200],"These":[201],"results":[202],"demonstrate":[203],"CheXmix":[205],"captures":[206],"information":[208],"broad":[211],"spectrum":[212],"X-ray":[215],"tasks.":[216],"code":[218],"at:":[220],"https://github.com/StanfordMIMI/CheXmix.":[221]},"counts_by_year":[],"updated_date":"2026-07-28T07:46:37.118299","created_date":"2026-04-29T00:00:00"}