{"id":"W4404838659","doi":"10.1016/j.neucom.2024.128998","title":"Depth-Wise Convolutions in Vision Transformers for efficient training on small datasets","year":2024,"lang":"en","type":"article","venue":"Neurocomputing","topic":"Advanced Neural Network Applications","field":"Computer Science","cited_by":44,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Computer science; Transformer; Artificial intelligence; Training (meteorology); Pattern recognition (psychology); Training set; Machine learning; Computer vision; Electrical engineering; Geography; Engineering; Voltage","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":[0.0006899593,0.001280086,0.0007193931,0.0005858436,0.0003340145,0.0009298743,0.002334603,0.0009240294,0.01161071],"category_scores_gemma":[0.002915435,0.0007354327,0.0009871488,0.0007589542,0.0006462957,0.002639241,0.001738333,0.002283347,0.004078876],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001094512,"about_ca_system_score_gemma":0.001450395,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00671635,"about_ca_topic_score_gemma":0.01202356,"domain_scores_codex":[0.999633,0.00005144108,0.00002871769,0.0001229215,0.00009388442,0.00007000771],"domain_scores_gemma":[0.9995539,0.0001421563,0.00003605087,0.0001594575,0.00007552017,0.00003281324],"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":[0.0004491878,0.0002200731,0.002338446,0.0004277529,0.0001890489,0.0002722951,0.000191759,0.2516509,0.04168427,0.03242075,0.02317076,0.6469848],"study_design_scores_gemma":[0.00003054526,0.00007754228,0.0003358872,0.00001644901,0.00002576327,0.0001338685,0.000021399,0.9663889,0.01377458,0.01439891,0.004783371,0.00001276577],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.02274422,0.0005522218,0.9597744,0.0002777141,0.00008888837,0.0001449875,0.0005975853,0.01179748,0.00402255],"genre_scores_gemma":[0.4606525,0.0006817614,0.5249214,0.0005443391,0.00007808813,0.0004560367,0.002944238,0.001282584,0.008439026],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01161071,"threshold_uncertainty_score":0.03884172,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04759880997046345,"score_gpt":0.3174072731901771,"score_spread":0.2698084632197136,"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."}}