{"id":"W7117137071","doi":"10.2139/ssrn.5964045","title":"Explainable AI: XVARS Outperforms SHAP in Terms of Computational Speed and Sampling Robustness","year":2025,"lang":"","type":"preprint","venue":"SSRN Electronic Journal","topic":"Explainable Artificial Intelligence (XAI)","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":false,"ca_institutions":"University of Saskatchewan","funders":"","keywords":"Robustness (evolution); Computation; Sampling (signal processing); Feature (linguistics); Sensitivity (control systems); Sample size determination; Explanatory power","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":["metaepi_narrow","research_integrity"],"consensus_categories":[],"category_scores_codex":[0.007848612,0.000926026,0.001439679,0.001645886,0.0007630516,0.0008384404,0.002987102,0.0006224521,0.00006632133],"category_scores_gemma":[0.0003729341,0.0009551705,0.0004069029,0.001550646,0.0004418305,0.001621132,0.002064492,0.008423271,0.00001388277],"about_ca_system_candidate":true,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.003423596,"about_ca_system_score_gemma":0.01025475,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0007741479,"about_ca_topic_score_gemma":0.001570418,"domain_scores_codex":[0.9887358,0.0004395603,0.002673709,0.001550424,0.001219448,0.005381081],"domain_scores_gemma":[0.9961142,0.0005470802,0.00137561,0.0008702463,0.0007832321,0.0003096573],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.0001687832,0.0003057716,0.004110759,0.0002893649,0.0002403078,0.00003080658,0.002053943,0.6905473,0.0001351819,0.2304974,0.00001185615,0.07160852],"study_design_scores_gemma":[0.0005645512,0.0003536414,0.0008677117,0.0008363196,0.00004878354,0.000417164,0.002270834,0.4846189,0.0004415456,0.5088802,0.00007380686,0.0006264945],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.2049295,0.001796883,0.7885631,0.001956181,0.001020874,0.00077704,0.00001408724,0.00004283799,0.0008994918],"genre_scores_gemma":[0.9822033,0.005523372,0.0104007,0.0002060324,0.0002774287,0.0000208653,0.00001891516,0.00004725176,0.001302127],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.7781624,"threshold_uncertainty_score":0.9992899,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03008068128982423,"score_gpt":0.3037943334211514,"score_spread":0.2737136521313272,"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."}}