{"id":"W4396738848","doi":"10.1016/j.earscirev.2024.104797","title":"An imperative for soil spectroscopic modelling is to think global but fit local with transfer learning","year":2024,"lang":"en","type":"article","venue":"Earth-Science Reviews","topic":"Soil Geostatistics and Mapping","field":"Environmental Science","cited_by":58,"is_retracted":false,"has_abstract":true,"ca_institutions":"Agriculture and Agri-Food Canada","funders":"Australian Research Council; Department of Industry, Science, Energy and Resources, Australian Government","keywords":"Soil carbon; Computer science; Spectral line; Local structure; Soil science; Property (philosophy); Environmental science; Artificial intelligence; Soil water; Chemistry; Physics","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":[0.0008700668,0.0001885138,0.0002468141,0.00003407106,0.0003855627,0.0002850372,0.0003071519,0.00003266166,0.0003859246],"category_scores_gemma":[0.0000302015,0.0001362059,0.00006330902,0.0007166889,0.000303452,0.0004603426,0.00004875537,0.0001512781,0.0005564149],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001476276,"about_ca_system_score_gemma":0.00008823183,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000219571,"about_ca_topic_score_gemma":0.0001862583,"domain_scores_codex":[0.9981455,0.0000423813,0.0002442847,0.000680467,0.0003871256,0.0005002855],"domain_scores_gemma":[0.9994485,0.00003355768,0.00002173483,0.0002307075,0.00001828304,0.000247171],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00006134648,0.00009627813,0.0008083867,0.0002989696,0.00002013718,0.00003462972,0.01471718,0.4444345,0.1041133,0.008886422,0.001273538,0.4252554],"study_design_scores_gemma":[0.0002583323,0.001425673,0.0005723895,0.0005249751,0.00005158378,0.0000263556,0.0005675525,0.7639614,0.02160094,0.001138943,0.2092215,0.0006503726],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.07979569,0.0008730362,0.916594,0.0002917092,0.0001250019,0.0006566533,0.0000166013,0.00005421676,0.001593082],"genre_scores_gemma":[0.8257647,0.001287501,0.1701048,0.001277707,0.0001270184,0.0001224242,0.000006101818,0.00002729961,0.001282463],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.7464892,"threshold_uncertainty_score":0.7151774,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02730851042707844,"score_gpt":0.2982563613040865,"score_spread":0.2709478508770081,"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."}}