{"id":"W2963511785","doi":"","title":"Learning seasonal phytoplankton communities with topic models","year":2017,"lang":"en","type":"article","venue":"arXiv (Cornell University)","topic":"Biomedical Text Mining and Ontologies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University","funders":"","keywords":"Interpretability; Probabilistic logic; Statistical model; Computer science; Set (abstract data type); Regression analysis; Regression; Probability distribution; Machine learning; Artificial intelligence; Econometrics; Mathematics; Statistics","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":[],"consensus_categories":[],"category_scores_codex":[0.00006345801,0.00009307682,0.00009046535,0.00001796565,0.00043009,0.0000365219,0.0003672225,0.0001007596,0.00002075295],"category_scores_gemma":[0.00002383257,0.00008438537,0.00003951273,0.00002386207,0.0003249047,0.000009055208,0.0001679885,0.0001316897,0.0000109465],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000008500763,"about_ca_system_score_gemma":0.00004224104,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001426838,"about_ca_topic_score_gemma":0.000148968,"domain_scores_codex":[0.9995493,0.00003954618,0.00003981702,0.0001767309,0.00003380509,0.000160795],"domain_scores_gemma":[0.9994254,0.0000127738,0.0000624567,0.0003962393,0.00004036752,0.00006276043],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"not_applicable","study_design_scores_codex":[0.002110838,0.0004191639,0.7970546,0.0001347436,0.0008528556,0.0009665191,0.001859098,0.07381078,0.00840364,0.06436757,0.002547547,0.04747261],"study_design_scores_gemma":[0.0205081,0.01397682,0.206717,0.0008075962,0.0006890681,0.0005584481,0.03207301,0.3147742,0.03918581,0.04107613,0.3240444,0.005589435],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9818352,0.00005729131,0.006059811,0.0000714495,0.00004229024,0.00003300394,0.000002793784,0.00002585594,0.01187234],"genre_scores_gemma":[0.9930874,0.000128575,0.0001491341,0.00006661309,0.00005512823,3.372048e-7,0.00001989717,0.00000725151,0.006485678],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.5903377,"threshold_uncertainty_score":0.3441136,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.07242095916689699,"score_gpt":0.1959279709067221,"score_spread":0.1235070117398251,"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."}}