{"id":"W7008845853","doi":"","title":"Deep networks training and generalization: insights from linearization","year":2023,"lang":"fr","type":"dissertation","venue":"Papyrus : Institutional Repository (Université de Montréal)","topic":"Stochastic Gradient Optimization Techniques","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Instituto de Ciencias del Mar y Limnología, Universidad Nacional Autónoma de México; Institute for Catastrophic Loss Reduction","keywords":"Linearization; Reprojection error; Counterweight","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow","sts"],"consensus_categories":[],"category_scores_codex":[0.0001294991,0.0005424622,0.000480164,0.0004985838,0.006384908,0.0003266533,0.0006227533,0.0007090761,0.00002034879],"category_scores_gemma":[0.0001353786,0.0006786599,0.0001697037,0.001093005,0.0002623698,0.0008688352,0.0003427992,0.0004151731,0.00002790789],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002038215,"about_ca_system_score_gemma":0.0009685268,"about_ca_topic_candidate":true,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01728208,"about_ca_topic_score_gemma":0.005404336,"domain_scores_codex":[0.9970114,0.0001995189,0.0005969801,0.001056462,0.0006831423,0.0004525036],"domain_scores_gemma":[0.9977543,0.0002559051,0.0005719143,0.0004897146,0.0005473486,0.0003808331],"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":[0.0001132778,0.0001110678,0.002329921,0.00007571297,0.0003387005,0.0009019015,0.1556586,0.5331857,0.001255193,0.2906577,0.0002397256,0.01513247],"study_design_scores_gemma":[0.0006129236,0.00009218702,0.01223336,0.0003702773,0.0002187869,0.0001538269,0.008672676,0.9711176,0.0008744759,0.002816003,0.002176347,0.0006615685],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.03333844,0.01717251,0.9417917,0.000263922,0.003309461,0.0005286809,0.0000199306,0.0006650071,0.002910397],"genre_scores_gemma":[0.8313843,0.005177314,0.1061233,0.0005124991,0.00150235,0.0001581637,0.006239609,0.0002075232,0.04869495],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.8356684,"threshold_uncertainty_score":0.9995664,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01261321955773689,"score_gpt":0.1853150738569515,"score_spread":0.1727018542992146,"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."}}