{"id":"W4389473428","doi":"10.16995/dscn.9669","title":"Alignement sémantique et manque de données : l’apport des modèles de langue. Le cas du latin et du grec","year":2023,"lang":"fr","type":"article","venue":"Digital Studies / Le champ numérique","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Humanities; Philosophy","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":true,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.01059505,0.0009657434,0.0009385755,0.004647959,0.001097565,0.009619041,0.001738003,0.001624002,0.004276781],"category_scores_gemma":[0.03401042,0.001084835,0.00248572,0.005393984,0.002758958,0.01186413,0.002828163,0.002780047,0.001831418],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002905598,"about_ca_system_score_gemma":0.003579283,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.02400355,"about_ca_topic_score_gemma":0.01918412,"domain_scores_codex":[0.9915991,0.003819709,0.0008081275,0.002056763,0.001554499,0.0001617732],"domain_scores_gemma":[0.9879858,0.007513611,0.0007716548,0.001982929,0.001566765,0.0001792148],"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":[0.0004928673,0.0001413132,0.01170873,0.001140935,0.0005541744,0.0004236633,0.007150198,0.04768804,0.007186423,0.3891909,0.01245773,0.5218651],"study_design_scores_gemma":[0.00004652251,0.0001445322,0.005631853,0.0005467238,0.0002725334,0.0006704054,0.002383844,0.5424685,0.005625994,0.2959583,0.1461008,0.0001499495],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.009575886,0.001580159,0.9792951,0.00227415,0.000158828,0.0001470053,0.001413332,0.002253273,0.003302142],"genre_scores_gemma":[0.195559,0.001747338,0.79189,0.0004629016,0.0001533363,0.0004178416,0.003545441,0.0008388924,0.005385367],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.02400355,"threshold_uncertainty_score":0.05603266,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05121940641463512,"score_gpt":0.3031767431746459,"score_spread":0.2519573367600108,"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."}}