{"id":"https://openalex.org/W7160721761","doi":"https://doi.org/10.1145/3805689.3806472","title":"Understanding Annotator Safety Policy with Interpretability","display_name":"Understanding Annotator Safety Policy with Interpretability","publication_year":2026,"publication_date":"2026-05-06","ids":{"openalex":"https://openalex.org/W7160721761","doi":"https://doi.org/10.1145/3805689.3806472"},"language":null,"primary_location":{"id":"pmh:oai:arXiv.org:2605.05329","is_oa":true,"landing_page_url":"https://arxiv.org/abs/2605.05329","pdf_url":"https://arxiv.org/pdf/2605.05329","source":{"id":"https://openalex.org/S4393918464","display_name":"ArXiv.org","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"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":"article","indexed_in":["arxiv"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2605.05329","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5068245182","display_name":"Alex Oesterling","orcid":"https://orcid.org/0000-0001-8546-0089"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Oesterling, Alex","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5086732355","display_name":"Donghao Ren","orcid":"https://orcid.org/0000-0001-8666-7241"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ren, Donghao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135779385","display_name":"Yannick Assogba","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Assogba, Yannick","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5006087928","display_name":"Dominik Moritz","orcid":"https://orcid.org/0000-0002-3110-1053"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Moritz, Dominik","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5061843649","display_name":"Sunnie S. Y. Kim","orcid":"https://orcid.org/0000-0002-8901-7233"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Kim, Sunnie S. Y.","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135757690","display_name":"Leon A. Gatys","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Gatys, Leon","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5034842879","display_name":"Fred Hohman","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Hohman, Fred","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":7,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":true,"cited_by_count":0,"citation_normalized_percentile":{"value":0.74844595,"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/T11704","display_name":"Mobile Crowdsensing and Crowdsourcing","score":0.5077000260353088,"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"}},"topics":[{"id":"https://openalex.org/T11704","display_name":"Mobile Crowdsensing and Crowdsourcing","score":0.5077000260353088,"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"}},{"id":"https://openalex.org/T10883","display_name":"Ethics and Social Impacts of AI","score":0.31700000166893005,"subfield":{"id":"https://openalex.org/subfields/3311","display_name":"Safety Research"},"field":{"id":"https://openalex.org/fields/33","display_name":"Social Sciences"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}},{"id":"https://openalex.org/T12026","display_name":"Explainable Artificial Intelligence (XAI)","score":0.03610000014305115,"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/interpretability","display_name":"Interpretability","score":0.8413000106811523},{"id":"https://openalex.org/keywords/ambiguity","display_name":"Ambiguity","score":0.8364999890327454},{"id":"https://openalex.org/keywords/counterfactual-thinking","display_name":"Counterfactual thinking","score":0.7998999953269958},{"id":"https://openalex.org/keywords/deliberation","display_name":"Deliberation","score":0.7181000113487244},{"id":"https://openalex.org/keywords/annotation","display_name":"Annotation","score":0.4756999909877777},{"id":"https://openalex.org/keywords/pluralism","display_name":"Pluralism (philosophy)","score":0.426800012588501}],"concepts":[{"id":"https://openalex.org/C2781067378","wikidata":"https://www.wikidata.org/wiki/Q17027399","display_name":"Interpretability","level":2,"score":0.8413000106811523},{"id":"https://openalex.org/C2780522230","wikidata":"https://www.wikidata.org/wiki/Q1140419","display_name":"Ambiguity","level":2,"score":0.8364999890327454},{"id":"https://openalex.org/C108650721","wikidata":"https://www.wikidata.org/wiki/Q1783253","display_name":"Counterfactual thinking","level":2,"score":0.7998999953269958},{"id":"https://openalex.org/C2776946740","wikidata":"https://www.wikidata.org/wiki/Q358652","display_name":"Deliberation","level":3,"score":0.7181000113487244},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6956999897956848},{"id":"https://openalex.org/C2776321320","wikidata":"https://www.wikidata.org/wiki/Q857525","display_name":"Annotation","level":2,"score":0.4756999909877777},{"id":"https://openalex.org/C49831778","wikidata":"https://www.wikidata.org/wiki/Q879021","display_name":"Pluralism (philosophy)","level":2,"score":0.426800012588501},{"id":"https://openalex.org/C2779530757","wikidata":"https://www.wikidata.org/wiki/Q1207505","display_name":"Quality (philosophy)","level":2,"score":0.3806000053882599},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3799999952316284},{"id":"https://openalex.org/C123587114","wikidata":"https://www.wikidata.org/wiki/Q2101508","display_name":"Policy analysis","level":2,"score":0.3361999988555908},{"id":"https://openalex.org/C112930515","wikidata":"https://www.wikidata.org/wiki/Q4389547","display_name":"Risk analysis (engineering)","level":1,"score":0.31709998846054077},{"id":"https://openalex.org/C56739046","wikidata":"https://www.wikidata.org/wiki/Q192060","display_name":"Knowledge management","level":1,"score":0.29159998893737793},{"id":"https://openalex.org/C2776291640","wikidata":"https://www.wikidata.org/wiki/Q2912517","display_name":"Value (mathematics)","level":2,"score":0.2831000089645386},{"id":"https://openalex.org/C206588197","wikidata":"https://www.wikidata.org/wiki/Q846574","display_name":"Reuse","level":2,"score":0.28299999237060547},{"id":"https://openalex.org/C2777503682","wikidata":"https://www.wikidata.org/wiki/Q5418327","display_name":"Evidence-based policy","level":3,"score":0.2639999985694885},{"id":"https://openalex.org/C2522767166","wikidata":"https://www.wikidata.org/wiki/Q2374463","display_name":"Data science","level":1,"score":0.2635999917984009},{"id":"https://openalex.org/C109986646","wikidata":"https://www.wikidata.org/wiki/Q546113","display_name":"Public policy","level":2,"score":0.26080000400543213},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.2596000134944916},{"id":"https://openalex.org/C539667460","wikidata":"https://www.wikidata.org/wiki/Q2414942","display_name":"Management science","level":1,"score":0.25360000133514404}],"mesh":[],"locations_count":1,"locations":[{"id":"pmh:oai:arXiv.org:2605.05329","is_oa":true,"landing_page_url":"https://arxiv.org/abs/2605.05329","pdf_url":"https://arxiv.org/pdf/2605.05329","source":{"id":"https://openalex.org/S4393918464","display_name":"ArXiv.org","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"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"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2605.05329","is_oa":true,"landing_page_url":"https://arxiv.org/abs/2605.05329","pdf_url":"https://arxiv.org/pdf/2605.05329","source":{"id":"https://openalex.org/S4393918464","display_name":"ArXiv.org","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"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":[{"id":"https://metadata.un.org/sdg/16","score":0.7729968428611755,"display_name":"Peace, Justice and strong institutions"}],"awards":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7160721761.pdf","grobid_xml":"https://content.openalex.org/works/W7160721761.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Safety":[0],"policies":[1,133],"define":[2],"what":[3],"constitutes":[4],"safe":[5],"and":[6,13,21,72,100,107,142,165,177,197,219],"unsafe":[7],"AI":[8],"outputs,":[9],"guiding":[10],"data":[11],"annotation":[12,17,98,146],"model":[14,153],"development.":[15],"However,":[16],"disagreement":[18],"is":[19,86,94],"pervasive":[20],"can":[22,101],"stem":[23],"from":[24,134],"multiple":[25],"sources":[26,57],"such":[27],"as":[28,110],"operational":[29,61],"failures":[30,62],"(annotators":[31],"misunderstand":[32],"or":[33,45],"misexecute":[34],"the":[35],"task),":[36],"policy":[37,70,156,168,187,222],"ambiguity":[38,67,188],"(policy":[39],"wording":[40],"leaves":[41],"room":[42],"for":[43,64,69,75,91,104],"interpretation),":[44],"value":[46,200],"pluralism":[47,73,201],"(different":[48],"annotators":[49,84,90,109,192],"hold":[50],"different":[51],"perspectives":[52],"on":[53],"safety).":[54],"Distinguishing":[55],"these":[56,213],"matters.":[58],"For":[59],"example,":[60],"call":[63],"quality":[65],"control,":[66],"calls":[68,74],"clarification,":[71],"deliberation":[76],"about":[77],"incorporating":[78],"diverse":[79],"perspectives.":[80],"Yet":[81],"understanding":[82],"why":[83],"disagree":[85],"difficult.":[87],"Directly":[88],"asking":[89],"their":[92],"reasoning":[93,112,140],"costly,":[95],"substantially":[96],"increasing":[97],"burden,":[99],"be":[102],"unreliable":[103],"both":[105],"human":[106,178],"LLM":[108,176],"self-reported":[111],"often":[113],"fails":[114],"to":[115,162,175],"reflect":[116],"actual":[117],"decision":[118],"processes.":[119],"We":[120,148],"introduce":[121],"Annotator":[122],"Policy":[123],"Models":[124],"(APMs),":[125],"interpretable":[126],"models":[127],"that":[128,150],"learn":[129],"annotators'":[130],"internal":[131],"safety":[132,155,194,207,221],"labeling":[135],"behavior":[136],"alone,":[137],"making":[138],"annotator":[139,154],"visible":[141],"comparable":[143],"without":[144],"additional":[145],"effort.":[147],"validate":[149],"APMs":[151,174],"accurately":[152],"(>80%":[157],"accuracy),":[158],"faithfully":[159],"predict":[160],"responses":[161],"counterfactual":[163],"edits,":[164],"recover":[166],"known":[167],"differences":[169,205],"in":[170,206],"controlled":[171],"settings.":[172],"Applying":[173],"annotations,":[179],"we":[180],"demonstrate":[181],"two":[182],"core":[183],"applications:":[184],"(1)":[185],"surfacing":[186,199],"by":[189,202],"revealing":[190],"how":[191],"interpret":[193],"instructions":[195],"differently,":[196],"(2)":[198],"uncovering":[203],"systematic":[204],"priorities":[208],"across":[209],"demographic":[210],"groups.":[211],"Together,":[212],"capabilities":[214],"support":[215],"more":[216],"targeted,":[217],"transparent,":[218],"inclusive":[220],"design.":[223]},"counts_by_year":[],"updated_date":"2026-06-11T09:08:48.828518","created_date":"2026-05-10T00:00:00"}