{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.002041246,0.002220388,0.0009696541,0.0009085096,0.0006047956,0.0007493718,0.002847737,0.001694981,0.006611063],"category_scores_gemma":[0.005679128,0.0008061657,0.001383599,0.0008437097,0.0006727958,0.001947726,0.001846071,0.003989958,0.003740175],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0009152465,"about_ca_system_score_gemma":0.001728652,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005765664,"about_ca_topic_score_gemma":0.01341588,"domain_scores_codex":[0.9987689,0.0004775939,0.00006295821,0.0004247822,0.0001636462,0.0001020957],"domain_scores_gemma":[0.9972414,0.001710401,0.000121878,0.0004083895,0.0004067832,0.0001112093],"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":[0.0007116435,0.0006355279,0.002519499,0.0006499621,0.0002169819,0.0003189027,0.0004080659,0.1879135,0.01841252,0.00378902,0.05739523,0.7270293],"study_design_scores_gemma":[0.0001008938,0.0001976609,0.0005566767,0.00003554204,0.00003247823,0.00008602056,0.00005301079,0.9753379,0.01136211,0.00514739,0.007062459,0.00002795066],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.03959532,0.001863025,0.8916691,0.0007007796,0.000489669,0.0006603799,0.002771723,0.05666924,0.005580693],"genre_scores_gemma":[0.3749739,0.0005667025,0.5920927,0.001452954,0.0002373887,0.002038168,0.01587742,0.002835659,0.009925126],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.006611063,"threshold_uncertainty_score":0.02211618,"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."}}