{"id":"W4309865903","doi":"10.1007/s00146-022-01591-z","title":"Toward safe AI","year":2022,"lang":"it","type":"article","venue":"AI & Society","topic":"Adversarial Robustness in Machine Learning","field":"Computer Science","cited_by":10,"is_retracted":false,"has_abstract":false,"ca_institutions":"Polytechnique Montréal","funders":"","keywords":"Computer science; Risk analysis (engineering); Normative; Artificial intelligence; Management science; Machine learning; Data science; Engineering","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":["metaepi_narrow","sts","research_integrity","insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.001370977,0.0003779054,0.0004131436,0.00004043796,0.001864097,0.0003456241,0.002401963,0.0001555969,0.002661505],"category_scores_gemma":[0.0001020973,0.0004417846,0.0006865939,0.001111425,0.0001930265,0.0006664963,0.003274822,0.002653113,0.0003112961],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006459429,"about_ca_system_score_gemma":0.0005676267,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000285737,"about_ca_topic_score_gemma":0.000002130442,"domain_scores_codex":[0.9959056,0.0005294502,0.0004679246,0.001011915,0.001208743,0.000876393],"domain_scores_gemma":[0.998152,0.0002336014,0.0002624186,0.001027267,0.0001256474,0.0001991063],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00003585426,0.0004060922,0.004639897,0.0001288299,0.0004315867,0.000166466,0.09739223,0.2583799,0.0003087307,0.065607,0.524071,0.04843243],"study_design_scores_gemma":[0.0008409003,0.0002026589,0.001216091,0.00001879647,0.00005941393,0.00005022523,0.002614253,0.5972673,0.00003752542,0.003088682,0.3939795,0.0006247098],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.00261509,0.001004593,0.9185439,0.06659374,0.005943627,0.0004041666,0.00003022027,0.0004277648,0.004436885],"genre_scores_gemma":[0.9340876,0.00008904008,0.0240142,0.03408524,0.0008811008,0.00005106124,0.00002122272,0.00006413524,0.006706402],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9314725,"threshold_uncertainty_score":0.9998034,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02187714235703819,"score_gpt":0.2800508867500709,"score_spread":0.2581737443930328,"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."}}