{"id":"W3101473090","doi":"10.26434/chemrxiv.11869692.v5","title":"RetroXpert: Decompose Retrosynthesis Prediction Like A Chemist","year":2020,"lang":"en","type":"article","venue":"ChemRxiv","topic":"Machine Learning in Materials Science","field":"Materials Science","cited_by":13,"is_retracted":false,"has_abstract":true,"ca_institutions":"NeuroDevNet","funders":"Cancer Prevention and Research Institute of Texas","keywords":"Retrosynthetic analysis; Interpretability; Synthon; Computer science; Artificial intelligence; Stepping stone; Margin (machine learning); Process (computing); Machine learning; Engineering; Chemistry; Programming language","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":["insufficient_payload"],"consensus_categories":["insufficient_payload"],"category_scores_codex":[0.0005258971,0.0002037474,0.0002632622,0.00002883053,0.0001976989,0.000193486,0.000630352,0.0001020523,0.004193068],"category_scores_gemma":[0.0008331103,0.0001863859,0.00007367849,0.0002770239,0.0001744043,0.0002523657,0.0001834819,0.0001575927,0.001307748],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00006172436,"about_ca_system_score_gemma":0.0000483394,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001742066,"about_ca_topic_score_gemma":7.19028e-7,"domain_scores_codex":[0.998143,0.0000793021,0.0003370343,0.0006419966,0.0004130618,0.0003856266],"domain_scores_gemma":[0.9990256,0.0000765181,0.0001557826,0.0004113062,0.00006562749,0.0002652056],"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":[0.00003931888,0.00002192379,0.002235464,0.00004773041,0.000002534229,0.000007363412,0.0003596344,0.0001048736,0.9915132,0.00003583485,0.00549435,0.000137756],"study_design_scores_gemma":[0.0002152116,0.00006990074,0.002406574,0.00002599554,0.00001263719,0.00001194903,0.00003595393,0.008438013,0.9752782,0.00005833044,0.01322845,0.0002187893],"study_design_candidate":"bench_or_experimental","study_design_consensus":"bench_or_experimental","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9899746,0.00006208986,0.001500268,0.002493877,0.0005898258,0.0001372166,0.000008727664,0.0005786252,0.00465478],"genre_scores_gemma":[0.9899092,0.00001037961,0.007991609,0.001240726,0.0004997672,0.00002376335,0.0000083918,0.0000306773,0.0002854647],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.01623502,"threshold_uncertainty_score":0.9994699,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01501172460227548,"score_gpt":0.2342369191532975,"score_spread":0.219225194551022,"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."}}