{"id":"https://openalex.org/W7153144115","doi":"https://doi.org/10.48550/arxiv.2604.08425","title":"Learning Who Disagrees: Demographic Importance Weighting for Modeling Annotator Distributions with DiADEM","display_name":"Learning Who Disagrees: Demographic Importance Weighting for Modeling Annotator Distributions with DiADEM","publication_year":2026,"publication_date":"2026-04-09","ids":{"openalex":"https://openalex.org/W7153144115","doi":"https://doi.org/10.48550/arxiv.2604.08425"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2604.08425","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.08425","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":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.2604.08425","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5120308315","display_name":"Samay U. Shetty","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Shetty, Samay U.","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133323732","display_name":"Tharindu Cyril Weerasooriya","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Weerasooriya, Tharindu Cyril","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5011457544","display_name":"Deepak Pandita","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Pandita, Deepak","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5133349741","display_name":"Christopher M. Homan","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Homan, Christopher M.","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/T12380","display_name":"Authorship Attribution and Profiling","score":0.16300000250339508,"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/T12380","display_name":"Authorship Attribution and Profiling","score":0.16300000250339508,"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":0.14630000293254852,"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/T11704","display_name":"Mobile Crowdsensing and Crowdsourcing","score":0.11129999905824661,"subfield":{"id":"https://openalex.org/subfields/1706","display_name":"Computer Science Applications"},"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/concatenation","display_name":"Concatenation (mathematics)","score":0.5939000248908997},{"id":"https://openalex.org/keywords/weighting","display_name":"Weighting","score":0.5929999947547913},{"id":"https://openalex.org/keywords/perspective","display_name":"Perspective (graphical)","score":0.5077000260353088},{"id":"https://openalex.org/keywords/coreference","display_name":"Coreference","score":0.39820000529289246},{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.3249000012874603},{"id":"https://openalex.org/keywords/semantics","display_name":"Semantics (computer science)","score":0.2930000126361847},{"id":"https://openalex.org/keywords/convolutional-neural-network","display_name":"Convolutional neural network","score":0.29170000553131104}],"concepts":[{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.7045999765396118},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7002000212669373},{"id":"https://openalex.org/C87619178","wikidata":"https://www.wikidata.org/wiki/Q126002","display_name":"Concatenation (mathematics)","level":2,"score":0.5939000248908997},{"id":"https://openalex.org/C183115368","wikidata":"https://www.wikidata.org/wiki/Q856577","display_name":"Weighting","level":2,"score":0.5929999947547913},{"id":"https://openalex.org/C12713177","wikidata":"https://www.wikidata.org/wiki/Q1900281","display_name":"Perspective (graphical)","level":2,"score":0.5077000260353088},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.4700999855995178},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.4180999994277954},{"id":"https://openalex.org/C28076734","wikidata":"https://www.wikidata.org/wiki/Q63087","display_name":"Coreference","level":3,"score":0.39820000529289246},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.3249000012874603},{"id":"https://openalex.org/C184337299","wikidata":"https://www.wikidata.org/wiki/Q1437428","display_name":"Semantics (computer science)","level":2,"score":0.2930000126361847},{"id":"https://openalex.org/C81363708","wikidata":"https://www.wikidata.org/wiki/Q17084460","display_name":"Convolutional neural network","level":2,"score":0.29170000553131104},{"id":"https://openalex.org/C189950617","wikidata":"https://www.wikidata.org/wiki/Q937228","display_name":"Property (philosophy)","level":2,"score":0.2800999879837036},{"id":"https://openalex.org/C108650721","wikidata":"https://www.wikidata.org/wiki/Q1783253","display_name":"Counterfactual thinking","level":2,"score":0.27790001034736633},{"id":"https://openalex.org/C77618280","wikidata":"https://www.wikidata.org/wiki/Q1155772","display_name":"Scheme (mathematics)","level":2,"score":0.27390000224113464},{"id":"https://openalex.org/C2780719617","wikidata":"https://www.wikidata.org/wiki/Q1030752","display_name":"Salient","level":2,"score":0.27320000529289246},{"id":"https://openalex.org/C2775936607","wikidata":"https://www.wikidata.org/wiki/Q466845","display_name":"Tracking (education)","level":2,"score":0.27000001072883606},{"id":"https://openalex.org/C86251818","wikidata":"https://www.wikidata.org/wiki/Q816754","display_name":"Benchmarking","level":2,"score":0.2660999894142151},{"id":"https://openalex.org/C2780110242","wikidata":"https://www.wikidata.org/wiki/Q5164387","display_name":"Construals","level":3,"score":0.26499998569488525},{"id":"https://openalex.org/C2776818064","wikidata":"https://www.wikidata.org/wiki/Q829903","display_name":"Agreement","level":2,"score":0.2565000057220459}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2604.08425","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.08425","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":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.2604.08425","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.08425","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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"sustainable_development_goals":[{"display_name":"Reduced inequalities","score":0.478705495595932,"id":"https://metadata.un.org/sdg/10"}],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"When":[0],"humans":[1],"label":[2,192],"subjective":[3],"content,":[4],"they":[5,191],"disagree,":[6],"and":[7,24,39,84,102,108,111,130,140,146,163],"that":[8,48,71,120,161,182,198],"disagreement":[9,118,151,175],"is":[10,112,193],"not":[11,188],"noise.":[12],"It":[13],"reflects":[14],"genuine":[15],"differences":[16],"in":[17],"perspective":[18],"shaped":[19],"by":[20,94],"annotators'":[21],"social":[22],"identities":[23],"lived":[25],"experiences.":[26],"Yet":[27],"standard":[28,145],"practice":[29],"still":[30],"flattens":[31],"these":[32],"judgments":[33],"into":[34],"a":[35,68,95,115],"single":[36],"majority":[37],"label,":[38],"recent":[40],"LLM-based":[41],"approaches":[42],"fare":[43],"no":[44],"better:":[45],"we":[46],"show":[47],"prompted":[49],"large":[50],"language":[51],"models,":[52],"even":[53],"with":[54,114],"chain-of-thought":[55],"reasoning,":[56],"fail":[57],"to":[58,200],"recover":[59],"the":[60,127,138,168],"structure":[61],"of":[62],"human":[63,203],"disagreement.":[64],"We":[65],"introduce":[66],"DiADEM,":[67],"neural":[69,141],"architecture":[70],"learns":[72],"\"how":[73],"much":[74],"each":[75],"demographic":[76,171],"axis":[77],"matters\"":[78],"for":[79,195],"predicting":[80],"who":[81,185],"will":[82],"disagree":[83],"on":[85,154],"what.":[86],"DiADEM":[87,134],"encodes":[88],"annotators":[89,186],"through":[90],"per-demographic":[91],"projections":[92],"governed":[93],"learned":[96,157],"importance":[97],"vector":[98],"$\\boldsymbol\u03b1$,":[99],"fuses":[100],"annotator":[101,174],"item":[103],"representations":[104],"via":[105],"complementary":[106],"concatenation":[107],"Hadamard":[109],"interactions,":[110],"trained":[113],"novel":[116],"item-level":[117],"loss":[119],"directly":[121],"penalizes":[122],"mispredicted":[123],"annotation":[124],"variance.":[125],"On":[126],"DICES":[128],"conversational-safety":[129],"VOICED":[131],"political-offense":[132],"benchmarks,":[133],"substantially":[135],"outperforms":[136],"both":[137,177],"LLM-as-a-judge":[139],"model":[142],"baselines":[143],"across":[144,176],"perspectivist":[147],"metrics,":[148],"achieving":[149],"strong":[150],"tracking":[152],"($r{=}0.75$":[153],"DICES).":[155],"The":[156],"$\\boldsymbol\u03b1$":[158],"weights":[159],"reveal":[160],"race":[162],"age":[164],"consistently":[165],"emerge":[166],"as":[167],"most":[169],"influential":[170],"factors":[172],"driving":[173],"datasets.":[178],"Our":[179],"results":[180],"demonstrate":[181],"explicitly":[183],"modeling":[184],"are":[187],"just":[189],"what":[190],"essential":[194],"NLP":[196],"systems":[197],"aim":[199],"faithfully":[201],"represent":[202],"interpretive":[204],"diversity.":[205]},"counts_by_year":[],"updated_date":"2026-07-28T07:46:37.118299","created_date":"2026-04-11T00:00:00"}