{"id":"W7132959475","doi":"","title":"Rethinking Clinical De-identification","year":2022,"lang":"","type":"dissertation","venue":"TSpace","topic":"Authorship Attribution and Profiling","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Vector Institute","keywords":"Focus (optics); Confidentiality; Task (project management); Scope (computer science); Bottleneck; Identity (music); Recall; Identification (biology)","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":["metaepi_narrow","sts","research_integrity","insufficient_payload"],"consensus_categories":["insufficient_payload"],"category_scores_codex":[0.01044881,0.0004543531,0.0005879099,0.00027989,0.001535677,0.0006734018,0.002091009,0.0007727713,0.002552234],"category_scores_gemma":[0.001482721,0.0005688794,0.0004952977,0.001219609,0.00009634177,0.0003835939,0.0004333481,0.002811573,0.0008245197],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0004315728,"about_ca_system_score_gemma":0.001452869,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001446943,"about_ca_topic_score_gemma":0.00004001944,"domain_scores_codex":[0.9928262,0.002312535,0.001547063,0.001453462,0.001096376,0.0007643853],"domain_scores_gemma":[0.9955922,0.0007604131,0.001567438,0.001362716,0.0003740634,0.0003431273],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"qualitative","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0004501416,0.0005237774,0.008902501,0.0007874275,0.0003906669,0.0002060982,0.5115778,0.0006443503,0.004871883,0.2703137,0.0105385,0.1907931],"study_design_scores_gemma":[0.002795384,0.001128951,0.07578677,0.001029876,0.0008810246,0.0001855845,0.05228933,0.4947485,0.02503872,0.09090215,0.2496936,0.005520174],"study_design_candidate":"qualitative","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.299042,0.003332887,0.6480891,0.006742815,0.02789122,0.001435485,0.00001911207,0.0008258966,0.01262153],"genre_scores_gemma":[0.8737128,0.0007535424,0.009439899,0.000829108,0.0007178343,0.0001108627,0.001278087,0.00007559443,0.1130823],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.6386492,"threshold_uncertainty_score":0.9999534,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1299195252492299,"score_gpt":0.4666195749250758,"score_spread":0.3367000496758459,"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."}}