{"id":"W4415195142","doi":"10.1609/aies.v8i3.36706","title":"Measuring What Matters: Connecting AI Ethics Evaluations to System Attributes, Hazards, and Harms","year":2025,"lang":"en","type":"article","venue":"Proceedings of the AAAI/ACM Conference on AI Ethics and Society","topic":"Ethics and Social Impacts of AI","field":"Social Sciences","cited_by":3,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University","funders":"","keywords":"Harm; Framing (construction); Situated; Set (abstract data type); Focus (optics)","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaresearch","sts","scholarly_communication","research_integrity"],"consensus_categories":[],"category_scores_codex":[0.01321759,0.0002413064,0.0004034304,0.00006719751,0.004221905,0.002091951,0.0007797069,0.0008665241,0.000009927913],"category_scores_gemma":[0.01462248,0.0002052652,0.0001765804,0.0005243576,0.0008432754,0.0008510474,0.0005732148,0.003354088,0.000002345561],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001760571,"about_ca_system_score_gemma":0.001355634,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001564996,"about_ca_topic_score_gemma":0.0003764757,"domain_scores_codex":[0.9972281,0.0002250309,0.0004183289,0.0004731442,0.001178279,0.0004771565],"domain_scores_gemma":[0.9932911,0.002135445,0.0002835838,0.0002263215,0.003826806,0.000236724],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"qualitative","study_design_scores_codex":[0.00001134474,0.00003307949,0.0009125832,0.000650144,0.0001196084,1.349587e-7,0.1994495,0.000004125061,0.002117375,0.7920111,0.003228383,0.001462685],"study_design_scores_gemma":[0.001297924,0.0002394067,0.003259362,0.0121528,0.0005166576,0.000003760421,0.7045016,0.001855611,0.005622452,0.2552301,0.01431711,0.001003219],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"commentary","genre_gemma":"empirical","genre_scores_codex":[0.165135,0.0004613374,0.0002327834,0.8251835,0.000689883,0.0008137401,0.00002532346,0.000117812,0.007340585],"genre_scores_gemma":[0.9631379,0.003725456,0.0007353325,0.03178259,0.000123224,0.00002479975,0.000001271184,0.00001780207,0.0004515753],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.798003,"threshold_uncertainty_score":0.9989452,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1888758223803678,"score_gpt":0.4198797978285312,"score_spread":0.2310039754481635,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}