{"id":"https://openalex.org/W4399794958","doi":"https://doi.org/10.48550/arxiv.2406.11230","title":"Multimodal Needle in a Haystack: Benchmarking Long-Context Capability of Multimodal Large Language Models","display_name":"Multimodal Needle in a Haystack: Benchmarking Long-Context Capability of Multimodal Large Language Models","publication_year":2024,"publication_date":"2024-06-17","ids":{"openalex":"https://openalex.org/W4399794958","doi":"https://doi.org/10.48550/arxiv.2406.11230"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2406.11230","is_oa":true,"landing_page_url":"http://arxiv.org/abs/2406.11230","pdf_url":"https://arxiv.org/pdf/2406.11230","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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"type":"preprint","indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2406.11230","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5040138884","display_name":"Hengyi Wang","orcid":"https://orcid.org/0000-0002-5325-2362"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Hengyi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101218444","display_name":"Haizhou Shi","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Shi, Haizhou","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5109867831","display_name":"Shiwei Tan","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Tan, Shiwei","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5111233245","display_name":"Weiyi Qin","orcid":"https://orcid.org/0000-0003-0846-649X"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Qin, Weiyi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100341750","display_name":"Wenyuan Wang","orcid":"https://orcid.org/0000-0002-8516-9168"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Wenyuan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5113257053","display_name":"Tunyu Zhang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhang, Tunyu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5089950211","display_name":"Akshay Nambi","orcid":"https://orcid.org/0000-0002-0921-4828"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Nambi, Akshay","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5024685185","display_name":"Tanuja Ganu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ganu, Tanuja","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5100446064","display_name":"Hao Wang","orcid":"https://orcid.org/0000-0001-9301-5989"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Hao","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":true,"cited_by_count":1,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.9817000031471252,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.9817000031471252,"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/T12031","display_name":"Speech and dialogue systems","score":0.9546999931335449,"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/T10028","display_name":"Topic Modeling","score":0.9531999826431274,"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/haystack","display_name":"Haystack","score":0.9799553155899048},{"id":"https://openalex.org/keywords/benchmarking","display_name":"Benchmarking","score":0.6912761330604553},{"id":"https://openalex.org/keywords/context","display_name":"Context (archaeology)","score":0.6791774034500122},{"id":"https://openalex.org/keywords/multimodal-therapy","display_name":"Multimodal therapy","score":0.5993838906288147},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.5249685645103455},{"id":"https://openalex.org/keywords/data-science","display_name":"Data science","score":0.3396263122558594},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.27626949548721313},{"id":"https://openalex.org/keywords/business","display_name":"Business","score":0.161239355802536},{"id":"https://openalex.org/keywords/psychology","display_name":"Psychology","score":0.14710626006126404},{"id":"https://openalex.org/keywords/geography","display_name":"Geography","score":0.14317122101783752}],"concepts":[{"id":"https://openalex.org/C13424479","wikidata":"https://www.wikidata.org/wiki/Q5687237","display_name":"Haystack","level":2,"score":0.9799553155899048},{"id":"https://openalex.org/C86251818","wikidata":"https://www.wikidata.org/wiki/Q816754","display_name":"Benchmarking","level":2,"score":0.6912761330604553},{"id":"https://openalex.org/C2779343474","wikidata":"https://www.wikidata.org/wiki/Q3109175","display_name":"Context (archaeology)","level":2,"score":0.6791774034500122},{"id":"https://openalex.org/C4441509","wikidata":"https://www.wikidata.org/wiki/Q6418787","display_name":"Multimodal therapy","level":2,"score":0.5993838906288147},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5249685645103455},{"id":"https://openalex.org/C2522767166","wikidata":"https://www.wikidata.org/wiki/Q2374463","display_name":"Data science","level":1,"score":0.3396263122558594},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.27626949548721313},{"id":"https://openalex.org/C144133560","wikidata":"https://www.wikidata.org/wiki/Q4830453","display_name":"Business","level":0,"score":0.161239355802536},{"id":"https://openalex.org/C15744967","wikidata":"https://www.wikidata.org/wiki/Q9418","display_name":"Psychology","level":0,"score":0.14710626006126404},{"id":"https://openalex.org/C205649164","wikidata":"https://www.wikidata.org/wiki/Q1071","display_name":"Geography","level":0,"score":0.14317122101783752},{"id":"https://openalex.org/C162853370","wikidata":"https://www.wikidata.org/wiki/Q39809","display_name":"Marketing","level":1,"score":0.0},{"id":"https://openalex.org/C166957645","wikidata":"https://www.wikidata.org/wiki/Q23498","display_name":"Archaeology","level":1,"score":0.0},{"id":"https://openalex.org/C542102704","wikidata":"https://www.wikidata.org/wiki/Q183257","display_name":"Psychotherapist","level":1,"score":0.0}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2406.11230","is_oa":true,"landing_page_url":"http://arxiv.org/abs/2406.11230","pdf_url":"https://arxiv.org/pdf/2406.11230","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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2406.11230","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2406.11230","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":"pmh:oai:arXiv.org:2406.11230","is_oa":true,"landing_page_url":"http://arxiv.org/abs/2406.11230","pdf_url":"https://arxiv.org/pdf/2406.11230","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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"awards":[{"id":"https://openalex.org/G202197749","display_name":"Computation for the Endless Frontier","funder_award_id":"1818253","funder_id":"https://openalex.org/F4320306076","funder_display_name":"National Science Foundation"},{"id":"https://openalex.org/G4002626585","display_name":"RI: Small: Enabling Interpretable AI via Bayesian Deep Learning","funder_award_id":"2127918","funder_id":"https://openalex.org/F4320306076","funder_display_name":"National Science Foundation"},{"id":"https://openalex.org/G6671297155","display_name":null,"funder_award_id":"CAREER","funder_id":"https://openalex.org/F4320306076","funder_display_name":"National Science Foundation"},{"id":"https://openalex.org/G6697622589","display_name":"CAREER: Robustify AI with Bayesian Deep Learning","funder_award_id":"2340125","funder_id":"https://openalex.org/F4320306076","funder_display_name":"National Science Foundation"}],"funders":[{"id":"https://openalex.org/F4320306076","display_name":"National Science Foundation","ror":"https://ror.org/021nxhr62"},{"id":"https://openalex.org/F4320308943","display_name":"Microsoft Research","ror":"https://ror.org/00d0nc645"},{"id":"https://openalex.org/F4320332161","display_name":"National Institutes of Health","ror":"https://ror.org/01cwqze88"}],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4399794958.pdf","grobid_xml":"https://content.openalex.org/works/W4399794958.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W4253878822","https://openalex.org/W1965563707","https://openalex.org/W4210692028","https://openalex.org/W1736550718","https://openalex.org/W2808729870","https://openalex.org/W2479343091","https://openalex.org/W2278064783","https://openalex.org/W3174858427","https://openalex.org/W1972480475","https://openalex.org/W275553111"],"abstract_inverted_index":{"Multimodal":[0],"Large":[1],"Language":[2],"Models":[3],"(MLLMs)":[4],"have":[5],"shown":[6],"significant":[7],"promise":[8],"in":[9,146,154,162],"various":[10],"applications,":[11],"leading":[12],"to":[13,44,58,69,85,191],"broad":[14],"interest":[15],"from":[16,151],"researchers":[17],"and":[18,65,101,116,134,181,188],"practitioners":[19],"alike.":[20],"However,":[21],"a":[22,67,87,92],"comprehensive":[23,166],"evaluation":[24,168],"of":[25,49,94,103,112,169],"their":[26,83],"long-context":[27,47,121,147,167],"capabilities":[28,48],"remains":[29],"underexplored.":[30],"To":[31],"address":[32],"these":[33],"gaps,":[34],"we":[35,54,127],"introduce":[36],"the":[37,46,61,163,175,185,193],"MultiModal":[38],"Needle-in-a-haystack":[39],"(MMNeedle)":[40],"benchmark,":[41,126],"specifically":[42],"designed":[43],"assess":[45],"MLLMs.":[50],"Besides":[51],"multi-image":[52],"input,":[53],"employ":[55],"image":[56,104,122],"stitching":[57],"further":[59],"increase":[60],"input":[62],"context":[63],"length,":[64],"develop":[66],"protocol":[68],"automatically":[70],"generate":[71],"labels":[72],"for":[73],"sub-image":[74,89],"level":[75],"retrieval.":[76],"Essentially,":[77],"MMNeedle":[78],"evaluates":[79],"MLLMs":[80,170],"by":[81],"stress-testing":[82],"capability":[84],"locate":[86],"target":[88],"(needle)":[90],"within":[91,120],"set":[93],"images":[95],"(haystack)":[96],"based":[97],"on":[98,174],"textual":[99],"instructions":[100,189],"descriptions":[102],"contents.":[105],"This":[106],"setup":[107],"necessitates":[108],"an":[109],"advanced":[110],"understanding":[111],"extensive":[113],"visual":[114],"contexts":[115],"effective":[117],"information":[118],"retrieval":[119],"inputs.":[123],"With":[124],"this":[125],"evaluate":[128],"state-of-the-art":[129],"MLLMs,":[130],"encompassing":[131],"both":[132],"API-based":[133,180],"open-source":[135,182],"models.":[136,183],"The":[137],"findings":[138],"reveal":[139],"that":[140],"GPT-4o":[141],"consistently":[142],"surpasses":[143],"other":[144],"models":[145],"scenarios,":[148],"but":[149],"suffers":[150],"hallucination":[152],"problems":[153],"negative":[155],"samples,":[156],"i.e.,":[157],"when":[158],"needles":[159],"are":[160,196],"not":[161],"haystacks.":[164],"Our":[165],"also":[171],"sheds":[172],"lights":[173],"considerable":[176],"performance":[177],"gap":[178],"between":[179],"All":[184],"code,":[186],"data,":[187],"required":[190],"reproduce":[192],"main":[194],"results":[195],"available":[197],"at":[198],"https://github.com/Wang-ML-Lab/multimodal-needle-in-a-haystack.":[199]},"counts_by_year":[{"year":2025,"cited_by_count":1}],"updated_date":"2026-08-26T07:47:46.906454","created_date":"2025-10-10T00:00:00"}