{"id":"W7106809904","doi":"10.48448/v1n5-c869","title":"Multilingual Language Model Pretraining using Machine-translated Data","year":2025,"lang":"","type":"other","venue":"Open MIND","topic":"","field":"","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University","funders":"","keywords":"Pipeline (software); Quality (philosophy); Welsh; Language model; Language proficiency; Lexical diversity; Commonsense reasoning; Order (exchange)","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":[0.002085373,0.002902694,0.001027497,0.001074309,0.001007668,0.001927532,0.002588708,0.001701199,0.01819647],"category_scores_gemma":[0.00959331,0.0010429,0.002210015,0.001208093,0.0007574456,0.003380806,0.002499297,0.005557955,0.01950428],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001076676,"about_ca_system_score_gemma":0.001981949,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01267986,"about_ca_topic_score_gemma":0.02907898,"domain_scores_codex":[0.9983678,0.0004362542,0.0001138661,0.0007483574,0.0001862121,0.000147451],"domain_scores_gemma":[0.9969869,0.001346407,0.00007487008,0.0008381745,0.0006436395,0.0001099668],"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":[0.0007213933,0.0009529811,0.008032497,0.001259839,0.0006498119,0.0007204068,0.0007996291,0.07935911,0.03040214,0.005118747,0.2526245,0.619359],"study_design_scores_gemma":[0.0004540038,0.0004502239,0.008580339,0.0004363773,0.0003572309,0.0006976248,0.0009779865,0.750605,0.08170827,0.01400813,0.1414694,0.0002554775],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.2009698,0.004865941,0.513343,0.002362343,0.003552211,0.001017971,0.05936009,0.1727941,0.04173457],"genre_scores_gemma":[0.3798438,0.000832999,0.3451127,0.001890233,0.0002511144,0.00140066,0.2405804,0.008690142,0.0213979],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.01819647,"threshold_uncertainty_score":0.06087321,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1916790069819072,"score_gpt":0.4434131125859522,"score_spread":0.251734105604045,"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."}}