{"id":"W4390091644","doi":"10.48550/arxiv.2312.12604","title":"An empirical study of testing machine learning in the wild","year":2023,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Software Testing and Debugging Techniques","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Computer science; Oracle; Software reliability testing; White-box testing; Regression testing; Software construction; Software engineering; Test strategy; System integration testing; Software system; Workflow; Software quality assurance; Software quality; Empirical research; Artificial intelligence; Machine learning; Software; Software development; Database; Programming language","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001138357,0.0002069032,0.000276993,0.0003541209,0.0001388535,0.00008729545,0.002507863,0.0001312123,0.00000101279],"category_scores_gemma":[0.000623495,0.0001856936,0.00006252708,0.001429416,0.00005371894,0.0001536166,0.001419032,0.0009651437,0.000006415946],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00005974745,"about_ca_system_score_gemma":0.00009812069,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002608972,"about_ca_topic_score_gemma":0.0002098344,"domain_scores_codex":[0.9980322,0.0005777553,0.0002264759,0.0007830656,0.0001396393,0.0002408913],"domain_scores_gemma":[0.9972644,0.001140834,0.0002350405,0.001201953,0.0001046956,0.00005300498],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.000005674866,0.0003524242,0.8607209,0.00002412205,0.00001360279,0.0003844552,0.002420295,0.134428,0.000002504784,0.0003783283,0.00015571,0.001113962],"study_design_scores_gemma":[0.0002186024,0.0004919872,0.09181459,0.0001077148,0.0000213141,0.000005270176,0.0002105373,0.8745031,0.000003026839,0.03239835,0.000008294884,0.0002172321],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.8276145,0.00001064276,0.1676373,0.00006090389,0.00009873674,0.0002818629,0.000001457058,0.004118913,0.0001757483],"genre_scores_gemma":[0.9935074,0.000004335327,0.00632679,0.00004827003,0.00002624403,0.000001926648,0.000003726036,0.00001572046,0.00006562707],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.7689063,"threshold_uncertainty_score":0.7572366,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2109586510133546,"score_gpt":0.2677698645024746,"score_spread":0.05681121348912002,"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."}}