{"id":"W3083386021","doi":"10.1101/2020.09.04.283929","title":"Self-Supervised Contrastive Learning of Protein Representations By Mutual Information Maximization","year":2020,"lang":"en","type":"preprint","venue":"bioRxiv (Cold Spring Harbor Laboratory)","topic":"Machine Learning in Bioinformatics","field":"Biochemistry, Genetics and Molecular Biology","cited_by":77,"is_retracted":false,"has_abstract":true,"ca_institutions":"Vector Institute; University of Toronto","funders":"Natural Sciences and Engineering Research Council of Canada; Canadian Institute for Advanced Research; Vector Institute; Microsoft Research","keywords":"Maximization; Computer science; Mutual information; Embedding; Artificial intelligence; Hyperparameter; Machine learning; Autoregressive model; Supervised learning; Contrast (vision); Mathematics; Artificial neural network","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":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.0003288266,0.0003857314,0.0003796164,0.0001288985,0.0001269699,0.0001223487,0.0004032066,0.0005280182,0.00002111493],"category_scores_gemma":[0.001099314,0.0004440311,0.0001295581,0.0002704829,0.00008666023,0.00004191892,0.0004689441,0.0006614967,0.00003136448],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00006211838,"about_ca_system_score_gemma":0.0004028775,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00002329782,"about_ca_topic_score_gemma":3.861695e-7,"domain_scores_codex":[0.997973,0.0002013676,0.0007548267,0.000422995,0.0003550734,0.000292764],"domain_scores_gemma":[0.9977142,0.00002334778,0.0008635235,0.0005964241,0.0006431088,0.000159408],"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":[0.0001131575,0.00009416912,0.005397572,0.001064921,0.0002878229,0.000001652327,0.0001294105,0.002642994,0.9888855,0.0003888594,0.0009813192,0.00001259633],"study_design_scores_gemma":[0.001619048,0.0004223996,0.007047886,0.0002505897,0.0001801703,3.875742e-8,0.00004813848,0.09015203,0.8701236,0.000002275559,0.02910033,0.001053513],"study_design_candidate":"bench_or_experimental","study_design_consensus":"bench_or_experimental","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.733438,0.0004317411,0.261691,0.0003797138,0.0003832433,0.002422902,0.000645529,0.0004064252,0.0002014752],"genre_scores_gemma":[0.9661594,0.0001226342,0.0331315,0.0001358504,0.0001591897,0.0001660457,0.00005521385,0.00006222323,0.000007964288],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.2327214,"threshold_uncertainty_score":0.9998012,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.006008598428195527,"score_gpt":0.2124216015829576,"score_spread":0.2064130031547621,"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."}}