{"id":"https://openalex.org/W7162128214","doi":"https://doi.org/10.48550/arxiv.2605.22168","title":"Measuring Cross-Modal Synergy: A Benchmark for VLM Explainability","display_name":"Measuring Cross-Modal Synergy: A Benchmark for VLM Explainability","publication_year":2026,"publication_date":"2026-05-21","ids":{"openalex":"https://openalex.org/W7162128214","doi":"https://doi.org/10.48550/arxiv.2605.22168"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2605.22168","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.22168","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":"Preprint"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2605.22168","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5063735145","display_name":"Jo\u00ebl Roman Ky","orcid":"https://orcid.org/0000-0001-8228-6847"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ky, Jo\u00ebl Roman","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5034416417","display_name":"Salah Ghamizi","orcid":"https://orcid.org/0000-0002-0738-8250"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ghamizi, Salah","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5000695937","display_name":"Maxime Cordy","orcid":"https://orcid.org/0000-0001-8312-1358"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Cordy, Maxime","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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.9934999942779541,"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":0.9934999942779541,"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":0.00139999995008111,"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/T10799","display_name":"Data Visualization and Analytics","score":0.0005000000237487257,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/benchmark","display_name":"Benchmark (surveying)","score":0.48240000009536743},{"id":"https://openalex.org/keywords/metric","display_name":"Metric (unit)","score":0.42089998722076416},{"id":"https://openalex.org/keywords/salience","display_name":"Salience (neuroscience)","score":0.3928000032901764},{"id":"https://openalex.org/keywords/probabilistic-logic","display_name":"Probabilistic logic","score":0.37709999084472656},{"id":"https://openalex.org/keywords/prior-probability","display_name":"Prior probability","score":0.37119999527931213},{"id":"https://openalex.org/keywords/visualization","display_name":"Visualization","score":0.3626999855041504},{"id":"https://openalex.org/keywords/visual-reasoning","display_name":"Visual reasoning","score":0.36169999837875366},{"id":"https://openalex.org/keywords/scalability","display_name":"Scalability","score":0.36070001125335693}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6406999826431274},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5593000054359436},{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.48240000009536743},{"id":"https://openalex.org/C176217482","wikidata":"https://www.wikidata.org/wiki/Q860554","display_name":"Metric (unit)","level":2,"score":0.42089998722076416},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.40070000290870667},{"id":"https://openalex.org/C108154423","wikidata":"https://www.wikidata.org/wiki/Q1469792","display_name":"Salience (neuroscience)","level":2,"score":0.3928000032901764},{"id":"https://openalex.org/C49937458","wikidata":"https://www.wikidata.org/wiki/Q2599292","display_name":"Probabilistic logic","level":2,"score":0.37709999084472656},{"id":"https://openalex.org/C177769412","wikidata":"https://www.wikidata.org/wiki/Q278090","display_name":"Prior probability","level":3,"score":0.37119999527931213},{"id":"https://openalex.org/C36464697","wikidata":"https://www.wikidata.org/wiki/Q451553","display_name":"Visualization","level":2,"score":0.3626999855041504},{"id":"https://openalex.org/C2777508537","wikidata":"https://www.wikidata.org/wiki/Q7936620","display_name":"Visual reasoning","level":2,"score":0.36169999837875366},{"id":"https://openalex.org/C48044578","wikidata":"https://www.wikidata.org/wiki/Q727490","display_name":"Scalability","level":2,"score":0.36070001125335693},{"id":"https://openalex.org/C184898388","wikidata":"https://www.wikidata.org/wiki/Q1435712","display_name":"Pairwise comparison","level":2,"score":0.3529999852180481},{"id":"https://openalex.org/C2993048729","wikidata":"https://www.wikidata.org/wiki/Q220821","display_name":"Visual methods","level":2,"score":0.3458000123500824},{"id":"https://openalex.org/C2780695315","wikidata":"https://www.wikidata.org/wiki/Q3799040","display_name":"Unobservable","level":2,"score":0.34139999747276306},{"id":"https://openalex.org/C185429906","wikidata":"https://www.wikidata.org/wiki/Q1130160","display_name":"Estimator","level":2,"score":0.321399986743927},{"id":"https://openalex.org/C2777146004","wikidata":"https://www.wikidata.org/wiki/Q14949826","display_name":"CLARITY","level":2,"score":0.30550000071525574},{"id":"https://openalex.org/C107673813","wikidata":"https://www.wikidata.org/wiki/Q812534","display_name":"Bayesian probability","level":2,"score":0.29899999499320984},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.27219998836517334},{"id":"https://openalex.org/C79581498","wikidata":"https://www.wikidata.org/wiki/Q1367530","display_name":"Suite","level":2,"score":0.2702000141143799},{"id":"https://openalex.org/C152124472","wikidata":"https://www.wikidata.org/wiki/Q1204361","display_name":"Redundancy (engineering)","level":2,"score":0.26750001311302185},{"id":"https://openalex.org/C2777055276","wikidata":"https://www.wikidata.org/wiki/Q7936580","display_name":"Visual approach","level":2,"score":0.2615000009536743},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.25519999861717224},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.25440001487731934},{"id":"https://openalex.org/C2778755073","wikidata":"https://www.wikidata.org/wiki/Q10858537","display_name":"Scale (ratio)","level":2,"score":0.2531000077724457},{"id":"https://openalex.org/C153083717","wikidata":"https://www.wikidata.org/wiki/Q6535263","display_name":"Leverage (statistics)","level":2,"score":0.25279998779296875}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2605.22168","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.22168","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":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2605.22168","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.22168","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":"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":{"Vision-Language":[0],"Models":[1],"(VLMs)":[2],"map":[3],"complex":[4],"visual":[5,52,68,141,156],"inputs":[6],"to":[7,50,169],"semantic":[8],"spaces,":[9],"but":[10],"interpreting":[11],"the":[12,92,99],"cross-modal":[13,46,152,159],"reasoning":[14,173],"of":[15],"VLMs":[16,43,137],"currently":[17],"relies":[18],"on":[19,140],"post-hoc":[20],"explainers":[21,134],"evaluated":[22],"via":[23],"unimodal":[24,58],"perturbation":[25],"metrics.":[26],"We":[27],"expose":[28],"a":[29,87,107,115,164],"limitation":[30],"in":[31,91,149,174],"this":[32,161],"paradigm:":[33],"because":[34],"multimodal":[35],"datasets":[36],"contain":[37],"language":[38],"priors":[39],"and":[40,69,128,143],"modality":[41],"biases,":[42],"frequently":[44],"exhibit":[45],"redundancy,":[47],"allowing":[48],"them":[49],"answer":[51],"queries":[53],"using":[54],"text":[55],"alone.":[56],"Consequently,":[57],"metrics":[59],"penalize":[60],"faithful":[61],"explainers,":[62],"triggering":[63],"an":[64],"evaluation":[65,166],"collapse":[66],"where":[67],"textual":[70],"rankings":[71],"fundamentally":[72],"contradict":[73],"each":[74],"other.":[75],"%(Kendall's":[76],"$\u03c4=":[77],"-0.06$).":[78],"To":[79],"resolve":[80],"this,":[81],"we":[82],"introduce":[83],"Synergistic":[84],"Faithfulness":[85],"($\\mathcal{F}_{syn}$),":[86],"scalable":[88],"metric":[89],"rooted":[90],"Shapley":[93],"Interaction":[94],"Index":[95],"that":[96,133],"strictly":[97],"isolates":[98],"joint":[100],"Harsanyi":[101],"dividend":[102],"between":[103],"modalities,":[104],"serving":[105],"as":[106],"highly":[108],"accurate":[109],"surrogate":[110],"($\u03c1=":[111],"0.92$)":[112],"while":[113],"achieving":[114],"$24\\times$":[116],"computational":[117],"speedup.":[118],"Evaluating":[119],"8":[120],"distinct":[121],"XAI":[122],"methods":[123,148],"across":[124],"3":[125,129],"VLM":[126,172],"architectures":[127],"benchmark":[130],"datasets,":[131],"reveals":[132],"proposed":[135],"for":[136],"heavily":[138],"over-index":[139],"salience":[142],"significantly":[144],"underperform":[145],"adapted":[146],"attention-based":[147],"capturing":[150],"true":[151],"synergy.":[153],"By":[154],"decoupling":[155],"plausibility":[157],"from":[158],"faithfulness,":[160],"work":[162],"provides":[163],"rigorous":[165],"framework":[167],"required":[168],"safely":[170],"audit":[171],"high-stakes":[175],"deployments.":[176]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-05-23T00:00:00"}