{"id":"W6968829696","doi":"10.5281/zenodo.4265632","title":"MeDAL: Medical Abbreviation Disambiguation Dataset for Natural Language Understanding Pretraining","year":2020,"lang":"en","type":"article","venue":"Zenodo (CERN European Organization for Nuclear Research)","topic":"Topic Modeling","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University","funders":"","keywords":"Download; Upload; Natural language; Text messaging; Natural (archaeology)","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":true,"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":[0.002098877,0.002356853,0.001219069,0.004099339,0.001490374,0.001343867,0.00301849,0.002982579,0.02802622],"category_scores_gemma":[0.009784666,0.0005243102,0.001605829,0.002397202,0.000809779,0.001771649,0.002645795,0.00261989,0.03803694],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001945946,"about_ca_system_score_gemma":0.003625447,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01341628,"about_ca_topic_score_gemma":0.02927416,"domain_scores_codex":[0.9975963,0.0005512353,0.0004334199,0.0007936361,0.0004493567,0.0001758694],"domain_scores_gemma":[0.9969335,0.001079522,0.0002290117,0.0006465875,0.0007285353,0.000382944],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","study_design_scores_codex":[0.0004105462,0.000147907,0.002264159,0.0014289,0.00008148145,0.0003441662,0.0001717269,0.0009887217,0.002376174,0.000997667,0.9595207,0.03126793],"study_design_scores_gemma":[0.001072395,0.0003601915,0.01838054,0.000594352,0.0001701015,0.002118808,0.000631833,0.01261984,0.009988935,0.004602951,0.9492639,0.0001962403],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"dataset","genre_gemma":"dataset","genre_scores_codex":[0.009205873,0.001308861,0.00623524,0.001505236,0.000466745,0.0007342814,0.9655787,0.008515803,0.006449197],"genre_scores_gemma":[0.006547213,0.0001665907,0.01000019,0.0005370414,0.00005076769,0.0006177048,0.9800149,0.0002569962,0.001808592],"genre_candidate":"dataset","genre_consensus":"dataset","teacher_disagreement_score":0.02802622,"threshold_uncertainty_score":0.09375703,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1124420204035989,"score_gpt":0.2983701788779148,"score_spread":0.1859281584743159,"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."}}