{"id":"W3089306998","doi":"10.48550/arxiv.2009.12452","title":"BET: A Backtranslation Approach for Easy Data Augmentation in Transformer-based Paraphrase Identification Context","year":2020,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Topic Modeling","field":"Computer Science","cited_by":3,"is_retracted":false,"has_abstract":true,"ca_institutions":"Polytechnique Montréal","funders":"","keywords":"Paraphrase; Generalizability theory; Computer science; Transformer; Artificial intelligence; Natural language processing; Deep learning; Training set; Machine learning; Statistics; 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":[0.002629083,0.001713052,0.0009283327,0.002467392,0.000806951,0.001625321,0.002628238,0.001512704,0.007574538],"category_scores_gemma":[0.01263201,0.000636996,0.001207167,0.002275325,0.0007138712,0.004610736,0.003793077,0.002793833,0.006035027],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006526918,"about_ca_system_score_gemma":0.001280594,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002936325,"about_ca_topic_score_gemma":0.007469925,"domain_scores_codex":[0.9984893,0.0005658449,0.0001737105,0.000408243,0.0002435413,0.0001193935],"domain_scores_gemma":[0.9963337,0.001197565,0.0001832573,0.00148724,0.000675606,0.0001226678],"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":[0.001064387,0.0005717666,0.005208217,0.0007738585,0.0002596045,0.0004176501,0.0007339466,0.0360469,0.03344732,0.008258345,0.04385154,0.8693664],"study_design_scores_gemma":[0.0002374497,0.0006432495,0.004195618,0.0001367556,0.0001523134,0.0007030452,0.0005246854,0.878075,0.05029235,0.02460198,0.04033012,0.0001074085],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.1558383,0.002553169,0.7576544,0.001325696,0.0006674359,0.0007103751,0.01349727,0.058075,0.009678421],"genre_scores_gemma":[0.470093,0.0006359522,0.4810188,0.0007562015,0.0002624365,0.001026631,0.03775042,0.001778718,0.006677708],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.007574538,"threshold_uncertainty_score":0.02533931,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2658474981201677,"score_gpt":0.2429260617664889,"score_spread":0.02292143635367885,"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."}}