{"id":"W2900457592","doi":"","title":"Deep Nets Don't Learn via Memorization","year":2017,"lang":"en","type":"article","venue":"PolyPublie (École Polytechnique de Montréal)","topic":"Stochastic Gradient Optimization Techniques","field":"Computer Science","cited_by":60,"is_retracted":false,"has_abstract":false,"ca_institutions":"McGill University; Polytechnique Montréal; Université de Montréal","funders":"","keywords":"Memorization; Computer science; Artificial intelligence; Mathematics education; Psychology","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","scholarly_communication"],"consensus_categories":[],"category_scores_codex":[0.0006618925,0.0003960781,0.0003649511,0.0004796709,0.001122146,0.001201544,0.003145362,0.0003105792,0.00004038241],"category_scores_gemma":[0.0005720302,0.0004176061,0.0001620698,0.000407892,0.0001631451,0.001732562,0.0009479875,0.0003788025,0.00004414974],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0003376894,"about_ca_system_score_gemma":0.000145686,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002491662,"about_ca_topic_score_gemma":0.0005842545,"domain_scores_codex":[0.9972494,0.0001078801,0.0005121181,0.0007697531,0.0005706617,0.0007902519],"domain_scores_gemma":[0.9956651,0.00008960735,0.0005944917,0.003024285,0.0002693907,0.0003570968],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00005557673,0.0004768826,0.029959,0.00004949791,0.0001009631,0.0001574151,0.0009766407,0.02702061,0.009118276,0.6630765,0.005377751,0.2636309],"study_design_scores_gemma":[0.0003897391,0.0001417766,0.0180308,0.00003679844,0.00001583447,0.0001112844,0.00001317362,0.9456688,0.01125917,0.02248584,0.001340358,0.0005064768],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.001225692,0.0002206302,0.9868866,0.005761264,0.0004439383,0.0007633637,0.000003732844,0.002337821,0.002356925],"genre_scores_gemma":[0.4969817,0.00007068028,0.5005857,0.001233404,0.0001111411,0.0002973701,0.00001274312,0.0000512009,0.0006560368],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.9186481,"threshold_uncertainty_score":0.9998353,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01166250967440062,"score_gpt":0.2350985713370607,"score_spread":0.2234360616626601,"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."}}