{"id":"W4415506496","doi":"10.1038/s41587-025-02859-7","title":"Democratizing protein language model training, sharing and collaboration","year":2025,"lang":"en","type":"article","venue":"Nature Biotechnology","topic":"Machine Learning in Bioinformatics","field":"Biochemistry, Genetics and Molecular Biology","cited_by":8,"is_retracted":false,"has_abstract":false,"ca_institutions":"University of Toronto","funders":"","keywords":"Language model; Training (meteorology); Language barrier; Training set; Deep learning","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":[0.0001454552,0.0001197869,0.0001136054,0.0001275281,0.00009273305,0.00003027304,0.0001902966,0.00106809,0.000002192259],"category_scores_gemma":[0.0003849879,0.0001160975,0.00002047183,0.0001996135,0.00007944402,0.000004783761,0.0002037667,0.0005070025,0.000001304622],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00001458375,"about_ca_system_score_gemma":0.00006933165,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000002360353,"about_ca_topic_score_gemma":0.00003966624,"domain_scores_codex":[0.999351,0.00001330943,0.0001489811,0.000259604,0.0000563085,0.0001707579],"domain_scores_gemma":[0.9995661,0.000004050236,0.00006608695,0.0003027376,0.00003986845,0.00002113158],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.00002854213,0.00001172029,0.000540852,0.00006081198,0.00003593065,0.000001946268,0.0002771327,0.0002323834,0.9509386,0.03585526,0.0009124941,0.01110428],"study_design_scores_gemma":[0.0009385261,0.0001998658,0.000197354,0.00009189494,0.00002907271,0.00003013189,0.001293921,0.1547364,0.8081143,0.003400742,0.03056334,0.0004045027],"study_design_candidate":"bench_or_experimental","study_design_consensus":"bench_or_experimental","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9752035,0.001760903,0.01488463,0.003482362,0.00008372529,0.000288371,0.000007432876,0.0001112771,0.004177809],"genre_scores_gemma":[0.9777456,0.00003941,0.02000502,0.000564631,0.00003406931,0.00001944705,0.00004588539,0.00001074879,0.001535229],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.154504,"threshold_uncertainty_score":0.823809,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.003583905128090461,"score_gpt":0.2719175898338231,"score_spread":0.2683336847057326,"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."}}