{"id":"W4285310604","doi":"10.18653/v1/2022.nlp4convai-1.5","title":"Data Augmentation for Intent Classification with Off-the-shelf Large Language Models","year":2022,"lang":"en","type":"article","venue":"","topic":"Topic Modeling","field":"Computer Science","cited_by":63,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Computer science; Classifier (UML); Training set; Task (project management); Labeled data; Language model; Machine learning; Artificial intelligence; Scarcity; Data quality; Data mining; Metric (unit)","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":[0.0004515552,0.00005243152,0.00004868476,0.0000309457,0.0002212246,0.00006942234,0.001131085,0.000008581645,0.00002299577],"category_scores_gemma":[0.000008185264,0.00003657772,0.00001341193,0.0001176694,0.000006794851,0.0004750896,0.0005857227,0.00006457011,0.000002492895],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0000508916,"about_ca_system_score_gemma":0.00004225267,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00003363287,"about_ca_topic_score_gemma":0.00006168064,"domain_scores_codex":[0.9991919,0.00003702258,0.0001172909,0.0003009741,0.0002219286,0.0001309064],"domain_scores_gemma":[0.9988881,0.00004373556,0.00005450299,0.0009618058,0.00003128632,0.00002053548],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00003039566,0.000136684,0.00007984357,0.00001609033,0.00003119038,0.000002347882,0.005905657,0.02209543,0.001060777,0.8407208,0.006538009,0.1233828],"study_design_scores_gemma":[0.0002876031,0.00003721052,0.00004079335,0.000001295675,0.000004908774,0.000003951097,0.00179877,0.9905517,0.0000899533,0.001678019,0.005448167,0.00005769255],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.004793735,0.00005587195,0.9905665,0.003265376,0.0001154299,0.0003321919,0.00004415745,0.00008756125,0.0007391837],"genre_scores_gemma":[0.8892365,0.000002344702,0.1083085,0.0009771603,0.00003287627,0.0001712349,0.0001791273,0.000006640164,0.001085553],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9684562,"threshold_uncertainty_score":0.2101857,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1125840059713915,"score_gpt":0.3134050345317351,"score_spread":0.2008210285603436,"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."}}