{"id":"W3101745556","doi":"10.18653/v1/2020.wmt-1.42","title":"Translating Similar Languages: Role of Mutual Intelligibility in Multilingual Transformers","year":2020,"lang":"en","type":"preprint","venue":"","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of British Columbia","funders":"Social Sciences and Humanities Research Council of Canada; Natural Sciences and Engineering Research Council of Canada; Compute Canada","keywords":"Computer science; Natural language processing; Jaccard index; Transformer; Artificial intelligence; Machine translation; Marathi; Intelligibility (philosophy); Speech recognition; Linguistics; Pattern recognition (psychology)","routes":{"ca_aff":true,"ca_fund":true,"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":[0.006329681,0.001419369,0.001122116,0.001501154,0.0007932867,0.003732674,0.001125377,0.0009860684,0.004248543],"category_scores_gemma":[0.01926636,0.0004998485,0.00108165,0.00139792,0.001384854,0.006310763,0.004102091,0.00213232,0.001725132],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007266115,"about_ca_system_score_gemma":0.001174991,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002441317,"about_ca_topic_score_gemma":0.003909979,"domain_scores_codex":[0.9953993,0.002281907,0.0003340282,0.001001423,0.0006786896,0.0003046305],"domain_scores_gemma":[0.9894297,0.005989201,0.000508444,0.002452702,0.001261053,0.0003590391],"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":[0.004789524,0.0009751294,0.04143477,0.001608193,0.001136391,0.001809911,0.0053641,0.2810828,0.08671723,0.03872942,0.008964225,0.5273884],"study_design_scores_gemma":[0.0001839742,0.001147092,0.007529361,0.0001168686,0.0005232646,0.001237335,0.001410861,0.8504403,0.07776456,0.04917173,0.0102894,0.0001852056],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.6354527,0.001501075,0.336398,0.001222679,0.0002240184,0.0001913369,0.001312721,0.005595424,0.01810202],"genre_scores_gemma":[0.9647027,0.0001880058,0.03202548,0.00008048877,0.00003778849,0.00003638852,0.0009387704,0.000490359,0.001499957],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.006329681,"threshold_uncertainty_score":0.03347498,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0230339289013167,"score_gpt":0.3216688303297862,"score_spread":0.2986349014284695,"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."}}