{"id":"https://openalex.org/W7137840458","doi":"https://doi.org/10.48550/arxiv.2603.14745","title":"CAMD: Coverage-Aware Multimodal Decoding for Efficient Reasoning of Multimodal Large Language Models","display_name":"CAMD: Coverage-Aware Multimodal Decoding for Efficient Reasoning of Multimodal Large Language Models","publication_year":2026,"publication_date":"2026-03-16","ids":{"openalex":"https://openalex.org/W7137840458","doi":"https://doi.org/10.48550/arxiv.2603.14745"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2603.14745","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.14745","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.2603.14745","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5129747263","display_name":"Huijie Guo","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Guo, Huijie","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5129727961","display_name":"Jingyao Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Jingyao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5129718171","display_name":"Lingyu Si","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Si, Lingyu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5129741212","display_name":"Jiahuan Zhou","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhou, Jiahuan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5129727938","display_name":"Changwen Zheng","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zheng, Changwen","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5129739126","display_name":"Wenwen Qiang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Qiang, Wenwen","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":6,"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":0.9060999751091003,"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":0.9060999751091003,"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/T10028","display_name":"Topic Modeling","score":0.01899999938905239,"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":0.011500000022351742,"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/security-token","display_name":"Security token","score":0.7105000019073486},{"id":"https://openalex.org/keywords/decoding-methods","display_name":"Decoding methods","score":0.6574000120162964},{"id":"https://openalex.org/keywords/benchmark","display_name":"Benchmark (surveying)","score":0.6370000243186951},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.5734000205993652},{"id":"https://openalex.org/keywords/residual","display_name":"Residual","score":0.4758000075817108},{"id":"https://openalex.org/keywords/reliability","display_name":"Reliability (semiconductor)","score":0.46939998865127563},{"id":"https://openalex.org/keywords/bayesian-inference","display_name":"Bayesian inference","score":0.45210000872612},{"id":"https://openalex.org/keywords/computation","display_name":"Computation","score":0.40389999747276306},{"id":"https://openalex.org/keywords/language-model","display_name":"Language model","score":0.38960000872612}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8130000233650208},{"id":"https://openalex.org/C48145219","wikidata":"https://www.wikidata.org/wiki/Q1335365","display_name":"Security token","level":2,"score":0.7105000019073486},{"id":"https://openalex.org/C57273362","wikidata":"https://www.wikidata.org/wiki/Q576722","display_name":"Decoding methods","level":2,"score":0.6574000120162964},{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.6370000243186951},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.5734000205993652},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5544999837875366},{"id":"https://openalex.org/C155512373","wikidata":"https://www.wikidata.org/wiki/Q287450","display_name":"Residual","level":2,"score":0.4758000075817108},{"id":"https://openalex.org/C43214815","wikidata":"https://www.wikidata.org/wiki/Q7310987","display_name":"Reliability (semiconductor)","level":3,"score":0.46939998865127563},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.46720001101493835},{"id":"https://openalex.org/C160234255","wikidata":"https://www.wikidata.org/wiki/Q812535","display_name":"Bayesian inference","level":3,"score":0.45210000872612},{"id":"https://openalex.org/C45374587","wikidata":"https://www.wikidata.org/wiki/Q12525525","display_name":"Computation","level":2,"score":0.40389999747276306},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.38960000872612},{"id":"https://openalex.org/C2779304628","wikidata":"https://www.wikidata.org/wiki/Q3503480","display_name":"Face (sociological concept)","level":2,"score":0.37139999866485596},{"id":"https://openalex.org/C107673813","wikidata":"https://www.wikidata.org/wiki/Q812534","display_name":"Bayesian probability","level":2,"score":0.3594000041484833},{"id":"https://openalex.org/C126042441","wikidata":"https://www.wikidata.org/wiki/Q1324888","display_name":"Frame (networking)","level":2,"score":0.33719998598098755},{"id":"https://openalex.org/C140779682","wikidata":"https://www.wikidata.org/wiki/Q210868","display_name":"Sampling (signal processing)","level":3,"score":0.33489999175071716},{"id":"https://openalex.org/C184337299","wikidata":"https://www.wikidata.org/wiki/Q1437428","display_name":"Semantics (computer science)","level":2,"score":0.3203999996185303},{"id":"https://openalex.org/C2780910867","wikidata":"https://www.wikidata.org/wiki/Q1952416","display_name":"Multimodality","level":2,"score":0.3075000047683716},{"id":"https://openalex.org/C81917197","wikidata":"https://www.wikidata.org/wiki/Q628760","display_name":"Selection (genetic algorithm)","level":2,"score":0.30090001225471497},{"id":"https://openalex.org/C125411270","wikidata":"https://www.wikidata.org/wiki/Q18653","display_name":"Encoding (memory)","level":2,"score":0.296999990940094},{"id":"https://openalex.org/C2776760102","wikidata":"https://www.wikidata.org/wiki/Q5139990","display_name":"Code (set theory)","level":3,"score":0.28999999165534973},{"id":"https://openalex.org/C49937458","wikidata":"https://www.wikidata.org/wiki/Q2599292","display_name":"Probabilistic logic","level":2,"score":0.2651999890804291},{"id":"https://openalex.org/C52740198","wikidata":"https://www.wikidata.org/wiki/Q1539564","display_name":"Importance sampling","level":3,"score":0.25690001249313354},{"id":"https://openalex.org/C120936955","wikidata":"https://www.wikidata.org/wiki/Q2155640","display_name":"Empirical research","level":2,"score":0.2556000053882599}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2603.14745","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.14745","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.2603.14745","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.14745","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":[],"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],"advances":[1],"in":[2],"Multimodal":[3,99],"Large":[4],"Language":[5],"Models":[6],"(MLLMs)":[7],"have":[8],"shown":[9],"impressive":[10],"reasoning":[11,73],"capabilities":[12],"across":[13],"vision-language":[14],"tasks,":[15],"yet":[16],"still":[17],"face":[18],"the":[19,88,143],"challenge":[20],"of":[21,82,147],"compute-difficulty":[22],"mismatch.":[23],"Through":[24],"empirical":[25],"analyses,":[26],"we":[27,53,96],"identify":[28],"that":[29,59,71,106],"existing":[30],"decoding":[31],"methods":[32],"may":[33],"waste":[34],"compute":[35],"on":[36,93,136],"easy":[37],"cases":[38],"while":[39],"underserving":[40],"hard":[41,83],"ones,":[42],"affecting":[43],"both":[44],"model":[45],"effectiveness":[46,144],"and":[47,65,121,128,140,145],"efficiency.":[48],"To":[49],"address":[50],"this":[51,94],"issue,":[52],"first":[54],"develop":[55],"a":[56,75,79,131],"theoretical":[57],"framework":[58],"links":[60],"sampling":[61],"coverage,":[62],"instance":[63],"difficulty,":[64],"residual":[66,89],"risk.":[67],"Our":[68],"analysis":[69],"reveals":[70],"multimodal":[72],"exhibits":[74],"heavy-tailed":[76],"difficulty":[77],"distribution;":[78],"small":[80],"subset":[81],"or":[84],"ambiguous":[85],"samples":[86],"dominates":[87],"failure":[90],"probability.":[91],"Based":[92],"insight,":[95],"propose":[97],"Coverage-Aware":[98],"Decoding":[100],"(CAMD),":[101],"an":[102],"adaptive":[103],"inference":[104],"mechanism":[105],"dynamically":[107],"allocates":[108],"computation":[109],"according":[110],"to":[111,125],"estimated":[112],"uncertainty.":[113],"CAMD":[114],"integrates":[115],"evidence-weighted":[116],"scoring,":[117],"posterior":[118],"coverage":[119],"estimation,":[120],"sequential":[122],"Bayesian":[123],"updating":[124],"balance":[126],"efficiency":[127],"reliability":[129],"under":[130],"limited":[132],"token":[133],"budget.":[134],"Experiments":[135],"various":[137],"benchmark":[138],"datasets":[139],"baselines":[141],"demonstrate":[142],"advantages":[146],"our":[148],"approach.":[149]},"counts_by_year":[],"updated_date":"2026-06-11T09:08:48.828518","created_date":"2026-03-18T00:00:00"}