{"id":"W4416996294","doi":"10.1007/s43681-025-00838-x","title":"Toward ethical AI through Bayesian uncertainty in neural question answering","year":2025,"lang":"en","type":"article","venue":"AI and Ethics","topic":"Topic Modeling","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":false,"ca_institutions":"HealthForceOntario","funders":"","keywords":"Interpretability; Bayesian probability; Bayesian inference; Inference; Artificial neural network; A priori and a posteriori; Calibration; Question answering","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":[],"consensus_categories":[],"category_scores_codex":[0.000759678,0.00008108604,0.0001095805,0.00007524602,0.00008699742,0.0001120725,0.0002404439,0.0003622615,0.000001650154],"category_scores_gemma":[0.0002517149,0.00007753559,0.00002116936,0.0002993747,0.00005291954,0.0002440159,0.0001766681,0.001809028,0.000001421913],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00002528056,"about_ca_system_score_gemma":0.0001079982,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0007728545,"about_ca_topic_score_gemma":0.0003458204,"domain_scores_codex":[0.999063,0.000155212,0.0001661025,0.000279063,0.0001570302,0.0001796298],"domain_scores_gemma":[0.9994536,0.000219825,0.0000177072,0.0002227463,0.00005374238,0.00003234747],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.000006652127,0.00001465266,0.002537976,0.0001100349,0.000005300845,0.00002077786,0.01051206,0.0131999,0.00006485349,0.9378046,0.0003647329,0.03535843],"study_design_scores_gemma":[0.0001619699,0.00001789116,0.00102294,0.0001044399,0.000002321567,0.000004462411,0.0000592819,0.8632774,0.00005840318,0.1328387,0.002359543,0.00009261957],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.006289828,0.0001751734,0.8224951,0.1697416,0.000441931,0.00005175277,2.454408e-7,0.00006658693,0.0007378046],"genre_scores_gemma":[0.9611788,0.0001317945,0.009670899,0.02892191,0.00003638389,0.000003230517,7.390117e-7,0.000002713222,0.00005348492],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.954889,"threshold_uncertainty_score":0.7859429,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05108168716002209,"score_gpt":0.3555435753322798,"score_spread":0.3044618881722577,"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."}}