{"id":"W2991989919","doi":"10.1109/tnnls.2019.2952322","title":"CHIP: Channel-Wise Disentangled Interpretation of Deep Convolutional Neural Networks","year":2019,"lang":"en","type":"article","venue":"IEEE Transactions on Neural Networks and Learning Systems","topic":"Explainable Artificial Intelligence (XAI)","field":"Computer Science","cited_by":24,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of British Columbia","funders":"Natural Sciences and Engineering Research Council of Canada; University of British Columbia","keywords":"Computer science; Discriminative model; Artificial intelligence; Convolutional neural network; Interpretation (philosophy); Machine learning; Pattern recognition (psychology); Class (philosophy); Regularization (linguistics); Channel (broadcasting); Deep learning","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":[0.0007194123,0.001259333,0.0005652619,0.0006202151,0.0002874885,0.001229411,0.001768919,0.001094721,0.00260096],"category_scores_gemma":[0.002732432,0.0004804289,0.0009259285,0.0005852327,0.001066961,0.001979062,0.00149045,0.002617232,0.000522684],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0009608174,"about_ca_system_score_gemma":0.001109276,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00366319,"about_ca_topic_score_gemma":0.005694195,"domain_scores_codex":[0.9995695,0.0001361567,0.00001617174,0.0001217593,0.0001036576,0.00005272511],"domain_scores_gemma":[0.999206,0.0002751174,0.0001312501,0.0001826131,0.000146703,0.00005822171],"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":[0.0004466696,0.0001174265,0.002825172,0.0002330005,0.0001555962,0.0003974988,0.0004082381,0.6281623,0.0228605,0.09869778,0.008810214,0.2368855],"study_design_scores_gemma":[0.00001087983,0.00003384703,0.0003322514,0.00001400849,0.00001715666,0.00004127818,0.00001877386,0.9578008,0.003090203,0.03728721,0.001340667,0.00001296603],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.01744038,0.0003294232,0.9782429,0.0003723456,0.00007158478,0.00003495611,0.0003341994,0.0009846621,0.002189605],"genre_scores_gemma":[0.7820228,0.0005779534,0.2091362,0.00045059,0.0001277488,0.0001488376,0.001259505,0.0004186331,0.005857819],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.00366319,"threshold_uncertainty_score":0.008701086,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01118582938150223,"score_gpt":0.2286773161765149,"score_spread":0.2174914867950127,"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."}}