{"id":"W7117729834","doi":"10.1186/s12874-025-02703-1","title":"Machine learning models explanations as interpretations of evidence: a theoretical framework of explainability and its implications on high-stakes biomedical decision-making","year":2025,"lang":"en","type":"article","venue":"BMC Medical Research Methodology","topic":"Explainable Artificial Intelligence (XAI)","field":"Computer Science","cited_by":3,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of British Columbia","funders":"European Commission","keywords":"Operationalization; Terminology; Soundness; Interpretation (philosophy); Interpretability; Raising (metalworking)","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":["metaresearch"],"consensus_categories":[],"category_scores_codex":[0.0231989,0.0001659926,0.0005526983,0.001060914,0.0003695892,0.00004998927,0.001732717,0.0003984435,0.0003522892],"category_scores_gemma":[0.4507954,0.0001270131,0.00010509,0.002223773,0.002073579,0.0003194731,0.001086042,0.001319094,0.00001363375],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001135029,"about_ca_system_score_gemma":0.001663475,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001669313,"about_ca_topic_score_gemma":0.00009066329,"domain_scores_codex":[0.9870662,0.008970654,0.001005225,0.0007684891,0.001606752,0.0005826573],"domain_scores_gemma":[0.6945377,0.3024474,0.0001988512,0.001013591,0.001380719,0.0004217743],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.0001543511,0.0001836516,0.0003377812,0.0001181917,0.00002390218,0.000006228933,0.001796502,0.0006540248,0.0003957548,0.8937401,0.00003828951,0.1025512],"study_design_scores_gemma":[0.0000784679,0.0003792445,0.0004318236,0.0009786666,0.000007962699,0.000008254433,0.0007406287,0.2734686,0.001722648,0.7220603,0.00005256033,0.00007081965],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.04846303,0.0008387243,0.9404163,0.009049136,0.0001652277,0.0004182834,0.00000710629,0.00005264828,0.0005895745],"genre_scores_gemma":[0.7317074,0.000420547,0.2675423,0.0001519523,0.00002282479,0.0001233808,0.000002039841,0.000006981054,0.00002254082],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.6832443,"threshold_uncertainty_score":0.8040323,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.4858301340420024,"score_gpt":0.5662124117095355,"score_spread":0.08038227766753309,"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."}}