{"id":"W4414419389","doi":"10.1038/s41592-025-02823-y","title":"Common to rare transfer learning (CORAL) enables inference and prediction for a quarter million rare Malagasy arthropods","year":2025,"lang":"en","type":"article","venue":"Nature Methods","topic":"Environmental DNA in Biodiversity Studies","field":"Environmental Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"Agriculture and Agri-Food Canada; University of Guelph","funders":"Academy of Finland; European Commission","keywords":"Inference; Context (archaeology); Scalability; Biodiversity; Rare species; Transfer of learning; Phylogenetic tree; DNA barcoding","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":[0.0004659951,0.0001946838,0.0002302676,0.00007413772,0.0004231745,0.00003626404,0.0001576896,0.0002665977,0.0001718798],"category_scores_gemma":[0.0001441985,0.000177917,0.00006632354,0.0002601733,0.000181957,0.0001754047,0.0002111303,0.0004258445,0.00002007842],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001521015,"about_ca_system_score_gemma":0.000002558513,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00004042708,"about_ca_topic_score_gemma":0.00005451752,"domain_scores_codex":[0.9987037,0.0002372676,0.0001693831,0.000455889,0.0001745587,0.0002591911],"domain_scores_gemma":[0.9994171,0.0002948782,0.00002544081,0.0001771229,0.000008814306,0.00007669901],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.0001860454,0.0000787371,0.8538737,0.00009431851,0.00006033758,0.000003331419,0.001895835,0.0008424208,0.02833572,0.000150283,0.01083749,0.1036417],"study_design_scores_gemma":[0.0008568026,0.0004715732,0.7887139,0.0001073046,0.0001033017,0.000003277214,0.001474502,0.0008171599,0.02395971,0.001096265,0.1820517,0.0003445655],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.8424144,0.001352065,0.14995,0.00199879,0.000550119,0.001141065,0.0001249692,0.0001261422,0.00234246],"genre_scores_gemma":[0.7535777,0.0002380728,0.2436635,0.001028192,0.00003284304,0.00007524587,0.00003798204,0.00001295922,0.0013335],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.1712142,"threshold_uncertainty_score":0.7255244,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0136211669560244,"score_gpt":0.3118259954445554,"score_spread":0.298204828488531,"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."}}