{"id":"W3119989665","doi":"","title":"ARBERT & MARBERT: Deep Bidirectional Transformers for Arabic.","year":2020,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":138,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of British Columbia","funders":"","keywords":"Computer science; Arabic; Transformer; Natural language processing; Inference; Language model; Task (project management); Artificial intelligence; Benchmark (surveying); Machine learning; Linguistics","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":[0.0009352138,0.002096088,0.0004821176,0.0009484062,0.0005628032,0.00160895,0.00198238,0.001302995,0.01074447],"category_scores_gemma":[0.003354115,0.0006694008,0.001188143,0.0006459666,0.000505908,0.003913511,0.002172231,0.002897189,0.01118284],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007751402,"about_ca_system_score_gemma":0.001182136,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005975824,"about_ca_topic_score_gemma":0.01087718,"domain_scores_codex":[0.9996809,0.00008169644,0.00002037745,0.0001091901,0.00006654281,0.00004139594],"domain_scores_gemma":[0.9994736,0.0001710915,0.000033421,0.0001533424,0.0001156278,0.00005293655],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.0006451418,0.0002147596,0.002102857,0.0004995551,0.0002045806,0.0004139716,0.0003945514,0.08453224,0.02197303,0.02650709,0.1631152,0.6993968],"study_design_scores_gemma":[0.00008601588,0.0001255467,0.0005698596,0.00009113724,0.00006036145,0.0002364019,0.0001563478,0.8874684,0.02700739,0.03774022,0.04640827,0.00004997968],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.03693563,0.003806376,0.8083197,0.002087309,0.00120496,0.0003106577,0.01064172,0.1136947,0.02299882],"genre_scores_gemma":[0.4275413,0.00192405,0.4894152,0.001430444,0.0001909156,0.0006462929,0.02900798,0.0044148,0.04542899],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.01074447,"threshold_uncertainty_score":0.03594387,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05721774270401291,"score_gpt":0.2069810579594347,"score_spread":0.1497633152554217,"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."}}