{"id":"W4385570792","doi":"10.18653/v1/2023.findings-acl.558","title":"MVP: Multi-task Supervised Pre-training for Natural Language Generation","year":2023,"lang":"en","type":"article","venue":"","topic":"Topic Modeling","field":"Computer Science","cited_by":18,"is_retracted":false,"has_abstract":true,"ca_institutions":"Université de Montréal","funders":"National Natural Science Foundation of China","keywords":"Computer science; Natural language generation; Task (project management); Artificial intelligence; Generality; Natural language processing; Language model; Natural language; Machine learning; Scale (ratio); Natural language understanding; Supervised learning; Artificial neural network","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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0002126703,0.00007115796,0.00007518693,0.00006642566,0.00008864069,0.0001059601,0.0003285174,0.00003363052,0.000007116476],"category_scores_gemma":[0.00005760027,0.00006175163,0.00004355398,0.0001752429,0.000005668273,0.000275999,0.00009952867,0.0000522794,0.00003043627],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00001560219,"about_ca_system_score_gemma":0.00002914458,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00003545499,"about_ca_topic_score_gemma":0.00006873537,"domain_scores_codex":[0.9992536,0.0000174527,0.0001256959,0.0002668366,0.00011909,0.0002173758],"domain_scores_gemma":[0.9995803,0.00005148723,0.00001911865,0.000278247,0.00003420089,0.00003663138],"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":[0.000005548886,0.00003280353,0.0002497062,0.0000458701,0.0000308837,0.00001714927,0.0396347,0.009583788,0.2839015,0.02599327,0.003911167,0.6365936],"study_design_scores_gemma":[0.000303149,0.00001092112,0.0004017798,0.000003539168,0.000001395426,0.000001580327,0.0001844554,0.9950871,0.003547546,0.00008097149,0.0002889227,0.00008865615],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.2016069,0.00005211864,0.796518,0.0005535588,0.000505008,0.0001750878,0.000001485089,0.0004668641,0.0001209548],"genre_scores_gemma":[0.7368831,0.000001176517,0.2600828,0.0002975669,0.0001787891,0.00002952161,0.00001835196,0.000005861573,0.002502811],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9855033,"threshold_uncertainty_score":0.2518159,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.08159341295343946,"score_gpt":0.3177525463533747,"score_spread":0.2361591333999353,"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."}}