{"id":"W2914453913","doi":"10.1101/526244","title":"Towards reliable named entity recognition in the biomedical domain","year":2019,"lang":"en","type":"preprint","venue":"bioRxiv (Cold Spring Harbor Laboratory)","topic":"Topic Modeling","field":"Computer Science","cited_by":9,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto","funders":"Compute Canada; National Institutes of Health; Nvidia","keywords":"Conditional random field; CRFS; Overfitting; Computer science; Artificial intelligence; Dropout (neural networks); Named-entity recognition; Transfer of learning; Machine learning; Sequence labeling; Deep learning; Natural language processing; Task (project management); Multi-task learning; Regularization (linguistics); Generalization; Artificial neural network","routes":{"ca_aff":true,"ca_fund":true,"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":[0.002607569,0.0004293302,0.0004845683,0.0003758212,0.00012266,0.000549511,0.002712827,0.0005988023,0.00002783175],"category_scores_gemma":[0.0002458861,0.000374493,0.000153749,0.0008772317,0.00009609706,0.0003694308,0.001281484,0.001285327,0.0002367754],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0003446267,"about_ca_system_score_gemma":0.0009140382,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0002529525,"about_ca_topic_score_gemma":0.000005499442,"domain_scores_codex":[0.996169,0.0003443901,0.0006554067,0.001282675,0.0008891302,0.0006593863],"domain_scores_gemma":[0.9966315,0.00009744061,0.0003269292,0.002530076,0.000248196,0.0001658602],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0003604317,0.009307214,0.07166775,0.01201217,0.001696715,0.00435024,0.004452195,0.004459881,0.6692891,0.1869401,0.03316818,0.002296146],"study_design_scores_gemma":[0.01072578,0.0007981625,0.3003608,0.00841312,0.0004080579,0.000001002414,0.0001186326,0.3753543,0.1110382,0.006551508,0.1742401,0.01199016],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.6263094,0.0004391379,0.3653409,0.002393097,0.003787197,0.00113444,0.00006328709,0.0004483277,0.00008424174],"genre_scores_gemma":[0.9142714,0.00009776306,0.08430573,0.0006764369,0.0004088055,0.0001949703,6.23396e-7,0.00004099532,0.000003286729],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.5582508,"threshold_uncertainty_score":0.9998707,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02133758801382152,"score_gpt":0.2280069111405203,"score_spread":0.2066693231266988,"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."}}