{"id":"W3107995456","doi":"10.48550/arxiv.2011.14578","title":"Where Should We Begin? A Low-Level Exploration of Weight Initialization Impact on Quantized Behaviour of Deep Neural Networks","year":2020,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Advanced Neural Network Applications","field":"Computer Science","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Initialization; Quantization (signal processing); Convolutional neural network; Computer science; Inference; Deep neural networks; Artificial intelligence; Deep learning; Artificial neural network; Algorithm","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"],"consensus_categories":[],"category_scores_codex":[0.0001094647,0.0004031652,0.0005601774,0.0002606658,0.0001049224,0.00004449226,0.001438977,0.0003134885,0.00001890467],"category_scores_gemma":[0.0000238349,0.000428242,0.0003274511,0.001295152,0.0001206336,0.000795842,0.0009205907,0.0005952015,0.000007289978],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001423327,"about_ca_system_score_gemma":0.00009011134,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00005158337,"about_ca_topic_score_gemma":0.00005545847,"domain_scores_codex":[0.9978771,0.0002116643,0.0004618561,0.0009706393,0.0001788676,0.0002999316],"domain_scores_gemma":[0.9972832,0.0001887241,0.0008954366,0.0011606,0.0002992004,0.0001729067],"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":[0.0001295208,0.0001417421,0.001278175,0.0000714535,0.00005074877,0.00003020526,0.000239454,0.9431937,0.0001352516,0.05317083,0.00009163438,0.001467327],"study_design_scores_gemma":[0.0005486856,0.000170231,0.001214812,0.0001408477,0.0000851868,0.00000151396,0.00003525452,0.9777029,0.0005646208,0.01918636,0.000009097543,0.0003405359],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.0674206,0.0001002314,0.9308385,0.0004037896,0.0002984887,0.0006268024,0.00004194271,0.0001710219,0.00009862606],"genre_scores_gemma":[0.9962956,0.0006199588,0.00275583,0.00006180644,0.0001017168,0.000005720661,0.0001016714,0.00003340982,0.00002430332],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.928875,"threshold_uncertainty_score":0.999817,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1694976685619838,"score_gpt":0.2618398569475958,"score_spread":0.092342188385612,"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."}}