{"id":"https://openalex.org/W7151067131","doi":"https://doi.org/10.48550/arxiv.2604.02923","title":"Council Mode: A Heterogeneous Multi-Agent Consensus Framework for Reducing LLM Hallucination and Bias","display_name":"Council Mode: A Heterogeneous Multi-Agent Consensus Framework for Reducing LLM Hallucination and Bias","publication_year":2026,"publication_date":"2026-04-03","ids":{"openalex":"https://openalex.org/W7151067131","doi":"https://doi.org/10.48550/arxiv.2604.02923"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2604.02923","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.02923","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":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.2604.02923","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5133027153","display_name":"Shuai Wu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wu, Shuai","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133020232","display_name":"Xue Li","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Xue","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101184173","display_name":"Yanna Feng","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Feng, Yanna","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133033072","display_name":"Yufang Li","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Yufang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133031090","display_name":"Zhijun Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Zhijun","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":null,"display_name":"Wang, Ran","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Ran","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/T10028","display_name":"Topic Modeling","score":0.36160001158714294,"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/T10028","display_name":"Topic Modeling","score":0.36160001158714294,"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/T11273","display_name":"Advanced Graph Neural Networks","score":0.08340000361204147,"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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.07029999792575836,"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/classifier","display_name":"Classifier (UML)","score":0.5012999773025513},{"id":"https://openalex.org/keywords/benchmark","display_name":"Benchmark (surveying)","score":0.42719998955726624},{"id":"https://openalex.org/keywords/triage","display_name":"Triage","score":0.4174000024795532},{"id":"https://openalex.org/keywords/matching","display_name":"Matching (statistics)","score":0.31779998540878296},{"id":"https://openalex.org/keywords/subject-matter-expert","display_name":"Subject-matter expert","score":0.3019999861717224},{"id":"https://openalex.org/keywords/pipeline","display_name":"Pipeline (software)","score":0.2913999855518341}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5509999990463257},{"id":"https://openalex.org/C95623464","wikidata":"https://www.wikidata.org/wiki/Q1096149","display_name":"Classifier (UML)","level":2,"score":0.5012999773025513},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.46129998564720154},{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.42719998955726624},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.4253000020980835},{"id":"https://openalex.org/C2777120189","wikidata":"https://www.wikidata.org/wiki/Q780067","display_name":"Triage","level":2,"score":0.4174000024795532},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.3472000062465668},{"id":"https://openalex.org/C2522767166","wikidata":"https://www.wikidata.org/wiki/Q2374463","display_name":"Data science","level":1,"score":0.34130001068115234},{"id":"https://openalex.org/C180747234","wikidata":"https://www.wikidata.org/wiki/Q23373","display_name":"Cognitive psychology","level":1,"score":0.3280999958515167},{"id":"https://openalex.org/C165064840","wikidata":"https://www.wikidata.org/wiki/Q1321061","display_name":"Matching (statistics)","level":2,"score":0.31779998540878296},{"id":"https://openalex.org/C105002631","wikidata":"https://www.wikidata.org/wiki/Q4833645","display_name":"Subject-matter expert","level":3,"score":0.3019999861717224},{"id":"https://openalex.org/C15744967","wikidata":"https://www.wikidata.org/wiki/Q9418","display_name":"Psychology","level":0,"score":0.30160000920295715},{"id":"https://openalex.org/C43521106","wikidata":"https://www.wikidata.org/wiki/Q2165493","display_name":"Pipeline (software)","level":2,"score":0.2913999855518341},{"id":"https://openalex.org/C166052673","wikidata":"https://www.wikidata.org/wiki/Q83021","display_name":"Empirical evidence","level":2,"score":0.2896000146865845},{"id":"https://openalex.org/C189708586","wikidata":"https://www.wikidata.org/wiki/Q1504425","display_name":"Systematic review","level":3,"score":0.28949999809265137},{"id":"https://openalex.org/C2775987171","wikidata":"https://www.wikidata.org/wiki/Q192797","display_name":"Elite","level":3,"score":0.2752000093460083},{"id":"https://openalex.org/C38652104","wikidata":"https://www.wikidata.org/wiki/Q3510521","display_name":"Computer security","level":1,"score":0.27410000562667847},{"id":"https://openalex.org/C196083921","wikidata":"https://www.wikidata.org/wiki/Q7915758","display_name":"Variance (accounting)","level":2,"score":0.2628999948501587},{"id":"https://openalex.org/C112930515","wikidata":"https://www.wikidata.org/wiki/Q4389547","display_name":"Risk analysis (engineering)","level":1,"score":0.2619999945163727},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.25699999928474426}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2604.02923","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.02923","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":"doi:10.48550/arxiv.2604.02923","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.02923","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":false,"raw_source_name":null,"raw_type":"Preprint"},"sustainable_development_goals":[{"display_name":"Quality Education","id":"https://metadata.un.org/sdg/4","score":0.6671879887580872}],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Large":[0],"Language":[1],"Models":[2],"(LLMs)":[3],"have":[4],"demonstrated":[5],"advanced":[6],"capabilities":[7],"but":[8],"often":[9],"suffer":[10],"from":[11],"factual":[12,219],"inaccuracies":[13],"(hallucinations)":[14],"and":[15,61,87,95,122,218],"systematic":[16],"biases.":[17],"These":[18,203],"issues,":[19],"sometimes":[20],"amplified":[21],"in":[22,59,114],"specific":[23],"architectures":[24],"like":[25],"Mixture-of-Experts":[26],"(MoE)":[27],"which":[28],"motivate":[29],"our":[30,99,135,168],"work,":[31],"pose":[32],"risks":[33],"for":[34,79,190,214],"reliable":[35],"deployment.":[36],"To":[37],"address":[38],"these":[39],"challenges,":[40],"we":[41],"propose":[42],"the":[43,106,130,141,155,179,194,199,216],"Council":[44,107,142],"Mode,":[45],"a":[46,66,88,110,118,123,145,151,174,182,211],"multi-agent":[47,208],"consensus":[48,68,209],"framework.":[49],"Our":[50],"approach":[51],"dispatches":[52],"queries":[53],"to":[54,129],"multiple":[55],"heterogeneous":[56],"frontier":[57],"LLMs":[58],"parallel":[60,82],"synthesizes":[62],"their":[63],"outputs":[64],"using":[65],"dedicated":[67],"model.":[69,133,158],"The":[70,159],"pipeline":[71],"consists":[72],"of":[73,148,196,221],"three":[74],"phases:":[75],"an":[76],"intelligent":[77],"triage":[78],"query":[80],"complexity,":[81],"generation":[83],"across":[84],"diverse":[85],"models,":[86],"structured":[89,207],"synthesis":[90],"that":[91,178,206],"identifies":[92],"agreement,":[93],"disagreement,":[94],"unique":[96],"findings.":[97],"In":[98],"evaluation,":[100],"conducted":[101],"under":[102,167],"controlled":[103],"no-web":[104],"settings,":[105],"Mode":[108,143],"achieved":[109,144],"41.7%":[111],"relative":[112],"reduction":[113],"hallucination":[115],"rates":[116],"on":[117,126],"1,200-sample":[119],"HaluEval":[120],"subset":[121],"7.5-point":[124],"improvement":[125,153],"TruthfulQA":[127],"compared":[128],"top-performing":[131],"individual":[132,157],"On":[134],"curated":[136],"MDR-500":[137],"multi-domain":[138],"reasoning":[139],"benchmark,":[140],"Quality":[146],"Score":[147],"95.4%,":[149],"representing":[150],"9.2-point":[152],"over":[154],"best":[156],"framework":[160,180],"also":[161],"exhibited":[162],"lower":[163],"measured":[164],"bias":[165],"variance":[166],"rubric-based":[169],"evaluation":[170],"protocol.":[171],"We":[172],"provide":[173],"cost-effectiveness":[175],"analysis":[176],"showing":[177],"incurs":[181],"4.2x":[183],"token-cost":[184],"overhead,":[185],"making":[186],"it":[187],"most":[188],"suitable":[189],"accuracy-prioritized":[191],"applications":[192],"where":[193],"cost":[195],"errors":[197],"exceeds":[198],"added":[200],"inference":[201],"cost.":[202],"findings":[204],"suggest":[205],"is":[210],"promising":[212],"direction":[213],"enhancing":[215],"reliability":[217],"grounding":[220],"LLM-generated":[222],"content.":[223]},"counts_by_year":[],"updated_date":"2026-07-28T07:46:37.118299","created_date":"2026-04-07T00:00:00"}