{"id":"W4412016444","doi":"10.1007/s10661-025-14293-y","title":"Unsupervised deep clustering as a tool for the identification of dark taxa in biomonitoring","year":2025,"lang":"en","type":"article","venue":"Environmental Monitoring and Assessment","topic":"Species Distribution and Climate Change","field":"Environmental Science","cited_by":3,"is_retracted":false,"has_abstract":false,"ca_institutions":"Dalhousie University","funders":"Science Fund of the Republic of Serbia","keywords":"Biomonitoring; Taxon; Identification (biology); Cluster analysis; Ecotoxicology; Environmental science; Ecology; Computer science; Biology; Artificial intelligence","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":[0.0002948748,0.0001101079,0.0001088579,0.00003226281,0.0001776456,0.00004218925,0.0001441768,0.0000404302,0.0002814162],"category_scores_gemma":[0.00001400929,0.0000938566,0.00004172804,0.0001148296,0.0001079291,0.0001070717,0.0001896067,0.00008143409,0.00001088908],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005858156,"about_ca_system_score_gemma":0.000004644992,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001254704,"about_ca_topic_score_gemma":0.00001000629,"domain_scores_codex":[0.9990865,0.00002517008,0.0002959439,0.0002315271,0.0001813762,0.0001795125],"domain_scores_gemma":[0.999606,0.00009701271,0.00008356211,0.0001849525,0.000001581956,0.00002693843],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.00002753867,0.0001558328,0.7150995,0.00003429992,0.00002159298,9.790774e-7,0.0003053444,0.0005215848,0.2395746,0.0001494814,0.00003085782,0.04407842],"study_design_scores_gemma":[0.0005347714,0.00003447237,0.9629343,0.00003185881,0.00002013683,8.790236e-7,0.004448237,0.001926737,0.02767972,0.00007760873,0.002207795,0.0001035045],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9954225,0.0002384672,0.002210215,0.0002191213,0.0004522754,0.0003836332,0.00001877408,0.00001243294,0.00104253],"genre_scores_gemma":[0.9987971,0.0003911013,0.0003362146,0.00001390436,0.00003057535,0.0001739944,0.000009936633,0.000007088379,0.0002400607],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.2478348,"threshold_uncertainty_score":0.3827361,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.016803895078171,"score_gpt":0.2984495858217892,"score_spread":0.2816456907436182,"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."}}