{"id":"W3099413717","doi":"10.18653/v1/2020.findings-emnlp.250","title":"Improving Word Embedding Factorization for Compression Using Distilled Nonlinear Neural Decomposition","year":2020,"lang":"en","type":"article","venue":"","topic":"Topic Modeling","field":"Computer Science","cited_by":8,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of British Columbia","funders":"","keywords":"Embedding; Computer science; Word embedding; Word (group theory); Matrix decomposition; Artificial intelligence; Machine translation; Translation (biology); Language model; Speech recognition; Natural language processing; Mathematics","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0007679442,0.001470141,0.000737073,0.0006495097,0.0003173911,0.000769046,0.0007331843,0.0007939136,0.004380003],"category_scores_gemma":[0.004691855,0.0003050321,0.0006819847,0.0008650372,0.0005399555,0.002493592,0.001205013,0.001694665,0.002735092],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0003998749,"about_ca_system_score_gemma":0.0008637967,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003616091,"about_ca_topic_score_gemma":0.006741147,"domain_scores_codex":[0.9995335,0.0001202746,0.00003456852,0.00010981,0.0001491367,0.00005261101],"domain_scores_gemma":[0.9986914,0.0005595107,0.00007681733,0.0003098048,0.0003119031,0.00005056016],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0004351291,0.0003417042,0.001180814,0.0002609464,0.0001042733,0.0002216216,0.0002726202,0.1831499,0.047698,0.01329196,0.01167051,0.7413725],"study_design_scores_gemma":[0.00003192975,0.00009179528,0.0002703043,0.00001546698,0.00002135667,0.0000884617,0.00005680613,0.9699733,0.02042335,0.006018216,0.002992647,0.00001646332],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.03919464,0.0004447464,0.9542438,0.0002170261,0.0001277479,0.00008267329,0.0003099106,0.003684625,0.001694892],"genre_scores_gemma":[0.3490361,0.0005381319,0.6387652,0.0002924325,0.0001364293,0.0002753628,0.003138742,0.0007073675,0.007110205],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.004380003,"threshold_uncertainty_score":0.01465261,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04846563712461783,"score_gpt":0.3185532309144585,"score_spread":0.2700875937898407,"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."}}