{"id":"W6929396995","doi":"10.48448/pggq-9g48","title":"Exploring the Effectiveness of Prompt Engineering for Legal Reasoning Tasks","year":2024,"lang":"en","type":"other","venue":"Underline Science Inc.","topic":"","field":"","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Thomson Reuters (Canada)","funders":"","keywords":"Task (project management); Textual entailment; Logical consequence; Natural language; Cluster analysis; Natural language understanding; Case-based reasoning; Automated reasoning","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":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.004852395,0.0004067668,0.0004652956,0.00103984,0.0001290918,0.0002172853,0.001143011,0.00009412732,0.00002953769],"category_scores_gemma":[0.001085603,0.0002905814,0.000133862,0.001789595,0.000697049,0.0003974846,0.0003214306,0.0004170823,0.0002963924],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0003103205,"about_ca_system_score_gemma":0.0004861983,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0003544287,"about_ca_topic_score_gemma":0.00006184981,"domain_scores_codex":[0.9972559,0.00007724511,0.0003174947,0.000842332,0.0008365149,0.0006705525],"domain_scores_gemma":[0.9983513,0.0004544403,0.000207692,0.0007235812,0.0001570363,0.0001059782],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"not_applicable","study_design_scores_codex":[0.0003490056,0.0003847911,0.0001411864,0.01527867,0.0009531588,0.00007265076,0.001531058,0.02174874,0.4266764,0.479016,0.03920273,0.01464562],"study_design_scores_gemma":[0.001854252,0.0007149021,0.0008835532,0.02750471,0.0007516645,0.000127545,0.0005787463,0.2033577,0.106361,0.001840071,0.653191,0.002834865],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"other","genre_gemma":"empirical","genre_scores_codex":[0.04407631,0.01135073,0.1349482,0.0004160867,0.03122285,0.02629726,0.002962106,0.010342,0.7383845],"genre_scores_gemma":[0.8825233,0.00004190241,0.03905773,0.00002197501,0.002372752,0.00253682,0.00009210151,0.004893725,0.06845972],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.838447,"threshold_uncertainty_score":0.9999546,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05129019214394856,"score_gpt":0.3005461493868248,"score_spread":0.2492559572428763,"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."}}