{"id":"W4389523920","doi":"10.18653/v1/2023.nllp-1.25","title":"A Comparative Study of Prompting Strategies for Legal Text Classification","year":2023,"lang":"en","type":"article","venue":"","topic":"Topic Modeling","field":"Computer Science","cited_by":7,"is_retracted":false,"has_abstract":true,"ca_institutions":"Vector Institute; Thomson Reuters (Canada)","funders":"","keywords":"Computer science; Domain (mathematical analysis); Natural language processing; Artificial intelligence; Mathematics","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":[0.007718106,0.001964519,0.001060906,0.002144696,0.0007829757,0.001940508,0.001623596,0.00166935,0.003297726],"category_scores_gemma":[0.0329393,0.0003961278,0.0008070781,0.001276197,0.0006230952,0.005454226,0.001395848,0.002945954,0.002574543],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001337113,"about_ca_system_score_gemma":0.002188195,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003191395,"about_ca_topic_score_gemma":0.005346255,"domain_scores_codex":[0.9950961,0.002560358,0.0003527191,0.001235095,0.0005270462,0.0002285971],"domain_scores_gemma":[0.976337,0.01839373,0.0007698622,0.001745769,0.001833865,0.0009198785],"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":[0.002644891,0.001661135,0.01890971,0.001179504,0.000210873,0.0002396476,0.001439591,0.05981448,0.01657554,0.003036494,0.01521996,0.8790681],"study_design_scores_gemma":[0.0003019568,0.00178609,0.006748467,0.0001544016,0.0002162844,0.0003439239,0.00163999,0.9367962,0.02948915,0.008749821,0.0136527,0.0001210781],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.5624619,0.01189494,0.360243,0.002689626,0.0009712367,0.001155132,0.002361867,0.04980153,0.008420862],"genre_scores_gemma":[0.7757332,0.001687168,0.2111253,0.0006070775,0.0002227466,0.0003816806,0.005425218,0.0007541657,0.004063506],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.007718106,"threshold_uncertainty_score":0.04081774,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2001642454534,"score_gpt":0.3770997134860705,"score_spread":0.1769354680326705,"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."}}