{"id":"W4413954097","doi":"10.1039/d5ta05426e","title":"Generalizable classification of crystal structure error types using graph attention networks","year":2025,"lang":"en","type":"article","venue":"Journal of Materials Chemistry A","topic":"Machine Learning in Materials Science","field":"Materials Science","cited_by":3,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Ottawa","funders":"Natural Sciences and Engineering Research Council of Canada; Total; Alliance de recherche numérique du Canada; Mitacs; Canada Foundation for Innovation; University of Ottawa","keywords":"Computer science; Artificial intelligence; Graph; Pattern recognition (psychology); Theoretical computer science","routes":{"ca_aff":true,"ca_fund":true,"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":["insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.001117978,0.0001690379,0.0004331395,0.00008421024,0.0001109317,0.0001573941,0.0004825021,0.0001564265,0.001526958],"category_scores_gemma":[0.0002128527,0.0001419032,0.00008271067,0.0002403333,0.0001527249,0.0002674574,0.00009499057,0.0001161217,0.000001851031],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00007046504,"about_ca_system_score_gemma":0.0001427761,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00002989894,"about_ca_topic_score_gemma":8.762271e-7,"domain_scores_codex":[0.99812,0.0001743996,0.0009242602,0.0002118518,0.0003357332,0.0002337908],"domain_scores_gemma":[0.9979725,0.00004118197,0.001330941,0.0002674754,0.000330005,0.00005792452],"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":[0.0001196699,0.00003157038,0.0004446327,0.0002529967,0.00001526612,0.000003269791,0.00002345824,0.04068904,0.9581071,0.00006965426,0.0002313474,0.00001201344],"study_design_scores_gemma":[0.0003209796,0.0000254613,0.002346862,0.0002529132,0.00006553273,0.00007620273,0.00004040399,0.003492112,0.992561,0.0005713936,0.000126286,0.0001208376],"study_design_candidate":"bench_or_experimental","study_design_consensus":"bench_or_experimental","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9861604,0.0001291066,0.01183804,0.00008652414,0.001524567,0.00007532784,0.00003505232,0.00001796603,0.0001330435],"genre_scores_gemma":[0.9900873,0.00001265421,0.009355886,0.00002613975,0.0003243611,0.000001245321,0.000009280423,0.00001254774,0.0001705647],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.03719693,"threshold_uncertainty_score":0.9993858,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01553909280166208,"score_gpt":0.2812368315697714,"score_spread":0.2656977387681093,"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."}}