{"id":"W4381683870","doi":"10.22214/ijraset.2023.53887","title":"Universal Language Model Fine-Tuning for Text Classification","year":2023,"lang":"en","type":"article","venue":"International Journal for Research in Applied Science and Engineering Technology","topic":"Topic Modeling","field":"Computer Science","cited_by":68,"is_retracted":false,"has_abstract":true,"ca_institutions":"Artificial Intelligence in Medicine (Canada)","funders":"","keywords":"Computer science; Artificial intelligence; Language model; Transfer of learning; Categorization; Classifier (UML); Natural language processing; Task (project management); Fine-tuning; Deep learning; Process (computing); Natural language; Machine learning; Programming language","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":[0.002039436,0.001505506,0.001037251,0.001770159,0.0007308641,0.001522535,0.001859555,0.001517553,0.004995124],"category_scores_gemma":[0.006761987,0.0003567901,0.001440148,0.001076719,0.0006756706,0.003757698,0.001772673,0.002603985,0.004330936],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001131691,"about_ca_system_score_gemma":0.001403253,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004091039,"about_ca_topic_score_gemma":0.005591126,"domain_scores_codex":[0.9983871,0.0003842884,0.0001511331,0.0006023858,0.0002717313,0.0002033782],"domain_scores_gemma":[0.997832,0.0008198584,0.0001588472,0.0005330809,0.000524269,0.0001320334],"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":[0.0003681529,0.0005069746,0.002699616,0.0003386658,0.000170732,0.0002286404,0.0002972118,0.07249903,0.05203712,0.006249766,0.03257485,0.8320293],"study_design_scores_gemma":[0.00003856816,0.00009605371,0.0007746771,0.00003929845,0.00004153472,0.0001638035,0.00008573369,0.9462022,0.02841314,0.01429306,0.00980463,0.00004727256],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.03512128,0.001680146,0.9388509,0.0005349472,0.0005727605,0.0002191437,0.0007701403,0.01881595,0.003434717],"genre_scores_gemma":[0.5322655,0.0006593855,0.4470532,0.001361064,0.000525725,0.0005945009,0.005453171,0.001694973,0.01039244],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.004995124,"threshold_uncertainty_score":0.0167104,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1047353107758972,"score_gpt":0.3965053319277147,"score_spread":0.2917700211518174,"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."}}