{"id":"W3215431778","doi":"10.18653/v1/2021.nllp-1.9","title":"JuriBERT: A Masked-Language Model Adaptation for French Legal Text","year":2021,"lang":"en","type":"preprint","venue":"","topic":"Artificial Intelligence in Law","field":"Social Sciences","cited_by":26,"is_retracted":false,"has_abstract":true,"ca_institutions":"Polytechnique Montréal","funders":"","keywords":"Computer science; Language model; Domain adaptation; Domain (mathematical analysis); Adaptation (eye); Set (abstract data type); Natural language processing; Focus (optics); Artificial intelligence; Programming language; Psychology; 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":[0.001695887,0.001645067,0.0005717534,0.0009780495,0.0006492578,0.001239242,0.001635965,0.001348626,0.01165412],"category_scores_gemma":[0.00759985,0.0006781769,0.001302552,0.0005644414,0.0004617221,0.002237363,0.001532275,0.002691793,0.006093651],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001252797,"about_ca_system_score_gemma":0.001840981,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.02850701,"about_ca_topic_score_gemma":0.03321678,"domain_scores_codex":[0.9991255,0.0003722862,0.00004730105,0.0002369265,0.0001370615,0.0000809226],"domain_scores_gemma":[0.9982643,0.0009720595,0.0000652272,0.0002666819,0.0003455954,0.00008624224],"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":[0.001441168,0.0005487701,0.006263041,0.001061452,0.0006965026,0.001390858,0.001109153,0.267421,0.02836616,0.01221831,0.2117139,0.4677697],"study_design_scores_gemma":[0.0001608776,0.000209219,0.00175242,0.00007943972,0.00008392308,0.0003466437,0.0001827184,0.9193884,0.01150965,0.007731459,0.05844866,0.000106679],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.1513877,0.002140566,0.6160814,0.002595629,0.00212831,0.001055784,0.02796816,0.175387,0.02125548],"genre_scores_gemma":[0.5186787,0.0009463336,0.3498741,0.001398686,0.000304699,0.001524277,0.07850716,0.01026876,0.03849731],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.02850701,"threshold_uncertainty_score":0.05668217,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.09036339189723698,"score_gpt":0.3887384578574484,"score_spread":0.2983750659602114,"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."}}