{"id":"W4376632853","doi":"10.48550/arxiv.2305.07507","title":"LeXFiles and LegalLAMA: Facilitating English Multinational Legal Language Model Development","year":2023,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Artificial Intelligence in Law","field":"Social Sciences","cited_by":4,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Fonds de recherche du Québec – Nature et technologies; Innovationsfonden","keywords":"Multinational corporation; Upstream (networking); Benchmark (surveying); Downstream (manufacturing); Computer science; Work (physics); Natural language processing; Knowledge management; Artificial intelligence; Business; Engineering; Marketing; Geography; Finance","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.006318785,0.001303548,0.0005859628,0.001560188,0.0006876431,0.002278393,0.003065106,0.0013818,0.01573289],"category_scores_gemma":[0.03579537,0.0007439903,0.0008033044,0.001147828,0.0007818558,0.006473888,0.003638795,0.0031495,0.006684072],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001763053,"about_ca_system_score_gemma":0.002621259,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.02257191,"about_ca_topic_score_gemma":0.03352917,"domain_scores_codex":[0.9969683,0.001560863,0.0002026558,0.0006174347,0.000490672,0.0001600707],"domain_scores_gemma":[0.9830155,0.01113965,0.000390457,0.003229475,0.00177887,0.0004460149],"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":[0.00112375,0.001818764,0.02917357,0.001521316,0.00039092,0.0008521142,0.003030244,0.1194952,0.01381879,0.01502947,0.2052319,0.608514],"study_design_scores_gemma":[0.0003418344,0.0005118245,0.01048543,0.0003811645,0.00009897444,0.0004616897,0.001887462,0.8321849,0.03027321,0.01314607,0.110055,0.000172493],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.5325162,0.00202673,0.1838703,0.005445755,0.0009869294,0.001287043,0.04528565,0.1865706,0.04201072],"genre_scores_gemma":[0.6174139,0.0006181512,0.2342539,0.001072855,0.0000881516,0.001174143,0.1272222,0.007790482,0.01036619],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.02257191,"threshold_uncertainty_score":0.05263174,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1709450320481825,"score_gpt":0.2690160056562647,"score_spread":0.09807097360808217,"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."}}