{"id":"W2986899672","doi":"10.1107/s2052520619013040","title":"Automated oxidation-state assignment for metal sites in coordination complexes in the Cambridge Structural Database","year":2019,"lang":"en","type":"article","venue":"Acta Crystallographica Section B Structural Science Crystal Engineering and Materials","topic":"Machine Learning in Materials Science","field":"Materials Science","cited_by":17,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Cambridge Crystallographic Data Centre; Engineering and Physical Sciences Research Council; McMaster University","keywords":"Scripting language; Workflow; Python (programming language); Computer science; Database; Oxidation state; Valence (chemistry); State (computer science); Data mining; Chemistry; Programming language; Metal","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.002596245,0.001311246,0.002081271,0.004907138,0.001070754,0.00242287,0.002642696,0.001026623,0.01503062],"category_scores_gemma":[0.006472532,0.0007069962,0.001145057,0.00381985,0.0004053212,0.001638738,0.001900076,0.001185618,0.0042583],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001505493,"about_ca_system_score_gemma":0.002873801,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005232129,"about_ca_topic_score_gemma":0.01411832,"domain_scores_codex":[0.9985878,0.0003002423,0.0001653675,0.0004402409,0.0003963962,0.0001099873],"domain_scores_gemma":[0.998604,0.0005368032,0.0001725472,0.0003853467,0.0002408168,0.0000605559],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.002927286,0.0004666933,0.05321303,0.01348686,0.001133552,0.001507092,0.002004968,0.01709522,0.09403863,0.0303448,0.4105596,0.3732223],"study_design_scores_gemma":[0.001387462,0.0005287167,0.05896236,0.001847308,0.001391437,0.0018543,0.001185898,0.2148837,0.1074431,0.03534877,0.574642,0.0005248919],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"dataset","genre_gemma":"empirical","genre_scores_codex":[0.3593382,0.01078098,0.1478591,0.001184168,0.0003960211,0.000963023,0.3718365,0.08222912,0.02541295],"genre_scores_gemma":[0.3222776,0.003725629,0.3026336,0.0002804353,0.00005879156,0.000771771,0.361438,0.005177387,0.003636745],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01503062,"threshold_uncertainty_score":0.05028242,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.00819728652259349,"score_gpt":0.2488270515487404,"score_spread":0.2406297650261469,"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."}}