{"id":"W4381713175","doi":"10.48550/arxiv.2306.11922","title":"No Wrong Turns: The Simple Geometry Of Neural Networks Optimization Paths","year":2023,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Stochastic Gradient Optimization Techniques","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Samsung; Natural Sciences and Engineering Research Council of Canada; Canadian Institute for Advanced Research","keywords":"Computer science; Artificial neural network; Initialization; Maxima and minima; Stochastic optimization; Artificial intelligence; Stochastic neural network; Simple (philosophy); Deep neural networks; Mathematical optimization; Machine learning; Algorithm; Mathematics; Recurrent neural network","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001399468,0.0004244604,0.0005607115,0.0005783246,0.000854291,0.001758194,0.0008175751,0.001105377,0.004716941],"category_scores_gemma":[0.01220011,0.0005168735,0.0005693958,0.0003232019,0.00267922,0.003740349,0.001668136,0.001894816,0.0006286441],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001187589,"about_ca_system_score_gemma":0.0007961122,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002015754,"about_ca_topic_score_gemma":0.001792121,"domain_scores_codex":[0.9993848,0.0002263135,0.00002383992,0.0001692624,0.0001236823,0.00007215521],"domain_scores_gemma":[0.9976488,0.001314306,0.0002909211,0.0003798193,0.0001945509,0.0001715782],"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":[0.0001804483,0.00004284426,0.002456483,0.00008162631,0.00003338662,0.0002352847,0.0003691812,0.3296854,0.002786874,0.6316382,0.003667489,0.02882277],"study_design_scores_gemma":[0.00002111419,0.0000550904,0.000922336,0.00003074556,0.000007573013,0.0000711004,0.00007430912,0.5133561,0.000931867,0.4817967,0.002712692,0.00002026717],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.2168417,0.0008797772,0.7617692,0.002780899,0.0001376258,0.00005977312,0.0003163253,0.0005965909,0.01661809],"genre_scores_gemma":[0.9363615,0.0004407174,0.05758496,0.0003361987,0.00004455236,0.0001051846,0.0001951896,0.0003398805,0.004591885],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.004716941,"threshold_uncertainty_score":0.01577973,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06147547643506778,"score_gpt":0.1947306989488869,"score_spread":0.1332552225138191,"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."}}