{"id":"https://openalex.org/W7134804254","doi":"https://doi.org/10.48550/arxiv.2603.08369","title":"M$^3$-ACE: Rectifying Visual Perception in Multimodal Math Reasoning via Multi-Agentic Context Engineering","display_name":"M$^3$-ACE: Rectifying Visual Perception in Multimodal Math Reasoning via Multi-Agentic Context Engineering","publication_year":2026,"publication_date":"2026-03-09","ids":{"openalex":"https://openalex.org/W7134804254","doi":"https://doi.org/10.48550/arxiv.2603.08369"},"language":null,"primary_location":{"id":"pmh:doi:10.48550/arxiv.2603.08369","is_oa":true,"landing_page_url":null,"pdf_url":null,"source":{"id":"https://openalex.org/S4406922384","display_name":"Open MIND","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Article"},"type":"article","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":null,"any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5128648517","display_name":"Peijin Xie","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xie, Peijin","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5128668690","display_name":"Zhen Xu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xu, Zhen","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5128677347","display_name":"Bingquan Liu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liu, Bingquan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5047681321","display_name":"Baoxun Wang","orcid":"https://orcid.org/0000-0001-8995-7543"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Baoxun","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":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.30145679,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"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":0.9384999871253967,"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":0.9384999871253967,"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/T10799","display_name":"Data Visualization and Analytics","score":0.0071000000461936,"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/T11273","display_name":"Advanced Graph Neural Networks","score":0.004600000102072954,"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/visual-reasoning","display_name":"Visual reasoning","score":0.7003999948501587},{"id":"https://openalex.org/keywords/perception","display_name":"Perception","score":0.6407999992370605},{"id":"https://openalex.org/keywords/context","display_name":"Context (archaeology)","score":0.6029999852180481},{"id":"https://openalex.org/keywords/benchmark","display_name":"Benchmark (surveying)","score":0.574999988079071},{"id":"https://openalex.org/keywords/visual-perception","display_name":"Visual perception","score":0.44859999418258667},{"id":"https://openalex.org/keywords/iterative-refinement","display_name":"Iterative refinement","score":0.39959999918937683}],"concepts":[{"id":"https://openalex.org/C2777508537","wikidata":"https://www.wikidata.org/wiki/Q7936620","display_name":"Visual reasoning","level":2,"score":0.7003999948501587},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.695900022983551},{"id":"https://openalex.org/C26760741","wikidata":"https://www.wikidata.org/wiki/Q160402","display_name":"Perception","level":2,"score":0.6407999992370605},{"id":"https://openalex.org/C2779343474","wikidata":"https://www.wikidata.org/wiki/Q3109175","display_name":"Context (archaeology)","level":2,"score":0.6029999852180481},{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.574999988079071},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5555999875068665},{"id":"https://openalex.org/C178253425","wikidata":"https://www.wikidata.org/wiki/Q162668","display_name":"Visual perception","level":3,"score":0.44859999418258667},{"id":"https://openalex.org/C107457646","wikidata":"https://www.wikidata.org/wiki/Q207434","display_name":"Human\u2013computer interaction","level":1,"score":0.41769999265670776},{"id":"https://openalex.org/C2779982483","wikidata":"https://www.wikidata.org/wiki/Q6094420","display_name":"Iterative refinement","level":2,"score":0.39959999918937683},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.36399999260902405},{"id":"https://openalex.org/C36464697","wikidata":"https://www.wikidata.org/wiki/Q451553","display_name":"Visualization","level":2,"score":0.3248000144958496},{"id":"https://openalex.org/C2780910867","wikidata":"https://www.wikidata.org/wiki/Q1952416","display_name":"Multimodality","level":2,"score":0.3001999855041504},{"id":"https://openalex.org/C165696696","wikidata":"https://www.wikidata.org/wiki/Q11287","display_name":"Exploit","level":2,"score":0.289900004863739},{"id":"https://openalex.org/C195344581","wikidata":"https://www.wikidata.org/wiki/Q2555318","display_name":"Automated reasoning","level":2,"score":0.28200000524520874},{"id":"https://openalex.org/C103057564","wikidata":"https://www.wikidata.org/wiki/Q4751139","display_name":"Analytic reasoning","level":3,"score":0.27869999408721924},{"id":"https://openalex.org/C2777055276","wikidata":"https://www.wikidata.org/wiki/Q7936580","display_name":"Visual approach","level":2,"score":0.26980000734329224}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:doi:10.48550/arxiv.2603.08369","is_oa":true,"landing_page_url":null,"pdf_url":null,"source":{"id":"https://openalex.org/S4406922384","display_name":"Open MIND","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Article"},{"id":"doi:10.48550/arxiv.2603.08369","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.08369","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":"pmh:doi:10.48550/arxiv.2603.08369","is_oa":true,"landing_page_url":null,"pdf_url":null,"source":{"id":"https://openalex.org/S4406922384","display_name":"Open MIND","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Article"},"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/4","display_name":"Quality Education","score":0.5158774256706238}],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Multimodal":[0],"large":[1],"language":[2],"models":[3,52],"have":[4],"recently":[5],"shown":[6],"promising":[7],"progress":[8],"in":[9,48,58,96],"visual":[10,26,42,94,120,185],"mathematical":[11,186],"reasoning.":[12,99],"However,":[13],"their":[14,59],"performance":[15,188],"is":[16],"often":[17],"limited":[18],"by":[19,112],"a":[20,86,115,151,167],"critical":[21],"yet":[22],"underexplored":[23],"bottleneck:":[24],"inaccurate":[25],"perception.":[27],"Through":[28],"systematic":[29],"analysis,":[30],"we":[31,83,145],"find":[32],"that":[33,154,170,181],"the":[34,130,200,218],"most":[35],"failures":[36],"originate":[37],"from":[38,157],"incorrect":[39],"or":[40,71],"incomplete":[41],"evidence":[43,121,156],"extraction":[44],"rather":[45],"than":[46],"deficiencies":[47],"reasoning":[49,111,187,227],"capability.":[50],"Moreover,":[51],"tend":[53],"to":[54,75,92,132],"remain":[55],"overly":[56],"confident":[57],"initial":[60],"perceptions,":[61],"making":[62],"standard":[63],"strategies":[64],"such":[65],"as":[66],"prompt":[67],"engineering,":[68],"multi-round":[69],"self-reflection,":[70],"posterior":[72],"guidance":[73],"insufficient":[74],"reliably":[76],"correct":[77],"errors.":[78],"To":[79,140],"address":[80],"this":[81],"limitation,":[82],"propose":[84],"M3-ACE,":[85],"multi-agentic":[87],"context":[88,117],"engineering":[89],"framework":[90],"designed":[91],"rectify":[93],"perception":[95,109],"multimodal":[97,226],"math":[98],"Instead":[100],"of":[101,220],"directly":[102],"aggregating":[103],"final":[104],"answers,":[105],"our":[106],"approach":[107],"decouples":[108],"and":[110,135,163,166,174,203,213],"dynamically":[113],"maintaining":[114],"shared":[116],"centered":[118],"on":[119,199,207],"lists.":[122],"Multiple":[123],"agents":[124,159],"collaboratively":[125],"contribute":[126],"complementary":[127],"observations,":[128],"enabling":[129],"system":[131],"expose":[133],"inconsistencies":[134],"recover":[136],"missing":[137],"perceptual":[138],"information.":[139],"support":[141],"stable":[142],"multi-turn":[143],"collaboration,":[144],"further":[146],"introduce":[147],"two":[148],"lightweight":[149],"tools:":[150],"Summary":[152],"Tool":[153,169],"organizes":[155],"different":[158],"into":[160],"consistent,":[161],"complementary,":[162],"conflicting":[164],"components,":[165],"Refine":[168],"filters":[171],"unreliable":[172],"samples":[173],"guides":[175],"iterative":[176],"correction.":[177],"Extensive":[178],"experiments":[179],"demonstrate":[180],"M3-ACE":[182],"substantially":[183],"improves":[184],"across":[189],"multiple":[190],"benchmarks.":[191],"Our":[192],"method":[193],"establishes":[194],"new":[195],"state-of-the-art":[196],"results":[197,216],"89.1":[198],"MathVision":[201],"benchmark":[202],"achieves":[204],"consistent":[205],"improvements":[206],"other":[208],"related":[209],"datasets,":[210],"including":[211],"MathVista":[212],"MathVerse.":[214],"These":[215],"highlight":[217],"importance":[219],"perception-centric":[221],"multi-agent":[222],"collaboration":[223],"for":[224],"advancing":[225],"systems.":[228]},"counts_by_year":[],"updated_date":"2026-07-15T18:14:33.161393","created_date":"2026-03-11T00:00:00"}