{"id":"W2787225861","doi":"10.1126/science.359.6377.725","title":"Artificial intelligence faces reproducibility crisis","year":2018,"lang":"en","type":"article","venue":"Science","topic":"Artificial Intelligence in Healthcare and Education","field":"Medicine","cited_by":573,"is_retracted":false,"has_abstract":true,"ca_institutions":"World Federation of Science Journalists","funders":"","keywords":"Reproducibility; Sensitivity (control systems); Code (set theory); Computer science; Artificial intelligence; Statistics; Mathematics; Programming language; 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":["insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.002887021,0.00008228635,0.0001217449,0.0001488165,0.000458883,0.00005769025,0.0002367834,0.00004506388,0.0004514254],"category_scores_gemma":[0.003061824,0.00006846294,0.00003504827,0.001303051,0.001469252,0.0002587502,0.00005983361,0.0001257412,0.001009449],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001196953,"about_ca_system_score_gemma":0.0006550552,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0003706295,"about_ca_topic_score_gemma":0.00009248532,"domain_scores_codex":[0.9978759,0.00002378796,0.0003668236,0.0009360036,0.000434005,0.0003634396],"domain_scores_gemma":[0.9977716,0.00005047197,0.00006367241,0.001145095,0.0007556133,0.0002135513],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.0001302794,0.000291163,0.03095437,0.00004371341,0.000004782866,0.000004278367,0.009452122,0.000005160221,0.05418393,0.004861831,0.001426477,0.8986419],"study_design_scores_gemma":[0.000002255928,0.0003659261,0.01096384,0.00002353195,0.000008820803,0.00001734777,0.00470017,0.0008249889,0.9448981,0.03617909,0.001917283,0.00009869927],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9807758,0.00007654585,0.003584289,0.01109922,0.001455079,0.0002556461,0.000001018461,0.00008211817,0.002670255],"genre_scores_gemma":[0.9955507,0.0000148341,0.002436172,0.001013913,0.000817394,0.000009602837,9.447173e-7,0.000005095866,0.0001513291],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.8985432,"threshold_uncertainty_score":0.9997684,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.3130665273602462,"score_gpt":0.4995019643729114,"score_spread":0.1864354370126652,"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."}}