{"id":"W2152808281","doi":"10.1109/tnn.2007.912312","title":"Adaptive Importance Sampling to Accelerate Training of a Neural Probabilistic Language Model","year":2008,"lang":"en","type":"article","venue":"IEEE Transactions on Neural Networks","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":233,"is_retracted":false,"has_abstract":true,"ca_institutions":"Université de Montréal","funders":"","keywords":"Computer science; Language model; Artificial neural network; Vocabulary; Probabilistic neural network; Speedup; Artificial intelligence; Probabilistic logic; Computation; Machine learning; Feedforward neural network; Sampling (signal processing); Statistical model; Training (meteorology); Backpropagation; Importance sampling; Time delay neural network; Algorithm; Statistics; Mathematics; Monte Carlo method","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":[],"consensus_categories":[],"category_scores_codex":[0.0001402725,0.00026208,0.0003148923,0.0001745258,0.0002485868,0.00005386811,0.0007629295,0.0001123633,0.000006295978],"category_scores_gemma":[0.00001301406,0.0002396311,0.0001327975,0.0007766355,0.00008085211,0.0005072049,0.000009942658,0.000557406,0.000001563153],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00006161699,"about_ca_system_score_gemma":0.00005933135,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00002553414,"about_ca_topic_score_gemma":0.00003843856,"domain_scores_codex":[0.9982831,0.00005684861,0.0003920132,0.0005271003,0.0002932349,0.0004477118],"domain_scores_gemma":[0.9989342,0.000141653,0.00014203,0.0005115244,0.0001167476,0.0001538397],"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":[0.00005792402,0.00005293938,0.000004013818,0.00001035082,0.0000114255,0.00003927288,0.002383558,0.9514742,0.002086079,0.0002551314,0.00003809709,0.043587],"study_design_scores_gemma":[0.0001672545,0.0001937094,0.00001599566,0.00005160719,0.00001037135,0.00007945515,0.00004113156,0.9937519,0.004827161,0.0006092986,0.000001216477,0.0002508653],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.08499976,0.000255237,0.9133865,0.0002065515,0.0002238197,0.0003455498,0.000009034253,0.000535349,0.0000382186],"genre_scores_gemma":[0.7971423,0.00000846205,0.2021116,0.0005631095,0.00004408604,0.00005544685,0.00000100484,0.00002278089,0.00005113383],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.7121426,"threshold_uncertainty_score":0.9771872,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0677840225644307,"score_gpt":0.2919126299227441,"score_spread":0.2241286073583134,"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."}}