{"id":"W4408549997","doi":"10.1029/2024wr037398","title":"On Robustness of the Explanatory Power of Machine Learning Models: Insights From a New Explainable AI Approach Using Sensitivity Analysis","year":2025,"lang":"en","type":"article","venue":"Water Resources Research","topic":"Explainable Artificial Intelligence (XAI)","field":"Computer Science","cited_by":23,"is_retracted":false,"has_abstract":true,"ca_institutions":"Global Institute for Water Security; University of Saskatchewan","funders":"Natural Sciences and Engineering Research Council of Canada; Suncor Energy Incorporated","keywords":"Robustness (evolution); Explanatory power; Sensitivity (control systems); Machine learning; Artificial intelligence; Computer science; Engineering; Epistemology; Chemistry","routes":{"ca_aff":true,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.01650715,0.001342347,0.001397071,0.003168324,0.0008280981,0.002899149,0.001843811,0.001421475,0.00282928],"category_scores_gemma":[0.07487557,0.0007250493,0.002585456,0.00116707,0.00359482,0.003004723,0.004227254,0.003703845,0.0001703439],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001701599,"about_ca_system_score_gemma":0.0009690865,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005153789,"about_ca_topic_score_gemma":0.001793705,"domain_scores_codex":[0.9944485,0.003867177,0.0002270847,0.0005521171,0.0006524342,0.0002525865],"domain_scores_gemma":[0.8776358,0.11299,0.003232523,0.003691562,0.001833856,0.0006162878],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00005101466,0.00005366417,0.004524559,0.000104575,0.0002297687,0.000189959,0.0002290384,0.8971295,0.0008064752,0.08758307,0.0005321204,0.008566303],"study_design_scores_gemma":[0.000003332904,0.00002059348,0.0005441722,0.00002347344,0.00001857693,0.00001854281,0.00002820164,0.9171452,0.0001766433,0.08178784,0.0002167313,0.00001668406],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1370893,0.001156929,0.848828,0.003374024,0.0001099875,0.0001020157,0.0002903765,0.0004350384,0.008614318],"genre_scores_gemma":[0.966815,0.000421913,0.03135096,0.0002979073,0.0001536826,0.0001010482,0.0001266584,0.0001136586,0.000619227],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01650715,"threshold_uncertainty_score":0.08729917,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.08346191017698525,"score_gpt":0.318397046788249,"score_spread":0.2349351366112638,"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."}}