{"id":"W4297785970","doi":"10.36227/techrxiv.21067438.v1","title":"A Trustworthy View on XAI Method Evaluation","year":2022,"lang":"en","type":"preprint","venue":"","topic":"Explainable Artificial Intelligence (XAI)","field":"Computer Science","cited_by":4,"is_retracted":false,"has_abstract":true,"ca_institutions":"Concordia University","funders":"","keywords":"Consistency (knowledge bases); Trustworthiness; Computer science; Centroid; Process (computing); Cluster analysis; Data mining; Order (exchange); Feature (linguistics); Artificial intelligence; Business; Computer security","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","insufficient_payload"],"consensus_categories":["insufficient_payload"],"category_scores_codex":[0.005349197,0.000346399,0.0004120989,0.0003268953,0.0002677702,0.0004163602,0.002624268,0.000182211,0.005759355],"category_scores_gemma":[0.0003507645,0.0003343758,0.0002426687,0.0006708386,0.00002562142,0.0002763452,0.003139986,0.0009248846,0.0008882516],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005233189,"about_ca_system_score_gemma":0.0006121055,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0004538961,"about_ca_topic_score_gemma":0.00008680179,"domain_scores_codex":[0.9949006,0.001266789,0.0005667877,0.001343454,0.001484356,0.0004380539],"domain_scores_gemma":[0.996626,0.0004222908,0.0002869821,0.002222741,0.0003167088,0.0001253409],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00000943009,0.0001454431,0.00001662585,0.00005521706,0.00004607017,0.00002163015,0.001078949,0.1206386,0.00003887537,0.2922277,0.00451586,0.5812057],"study_design_scores_gemma":[0.00006271531,0.0001361876,0.00006354191,0.00005273696,0.00003277276,0.0000051372,0.0001373236,0.7765979,0.003383482,0.1675773,0.05150981,0.0004411104],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.0006187751,0.0004186172,0.8882121,0.003074306,0.002318518,0.001200058,0.000004217596,0.0004141711,0.1037392],"genre_scores_gemma":[0.1138094,0.0003904926,0.8593702,0.009233121,0.0006882389,0.003256172,0.0001195301,0.0001131753,0.01301971],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.6559592,"threshold_uncertainty_score":0.9999108,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1202530959946061,"score_gpt":0.4160330938312373,"score_spread":0.2957799978366311,"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."}}