{"id":"W6939717743","doi":"10.6084/m9.figshare.28059556.v1","title":"Additional file 3 of CGRclust: Chaos Game Representation for twin contrastive clustering of unlabelled DNA sequences","year":2024,"lang":"en","type":"article","venue":"Figshare","topic":"Fractal and DNA sequence analysis","field":"Biochemistry, Genetics and Molecular Biology","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Western University; University of Waterloo","funders":"","keywords":"Cluster analysis; Representation (politics); CHAOS (operating system); Pattern recognition (psychology); Feature (linguistics)","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":[],"category_scores_codex":[0.00001674081,0.00007498375,0.0001190708,0.00004129773,0.00001948273,0.00001105679,0.00008450248,0.00007079857,0.7266422],"category_scores_gemma":[0.001124672,0.00006848831,0.0001330148,0.0001120948,0.00001867788,0.000008815473,0.00004501958,0.00003078124,0.00005574404],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000007587137,"about_ca_system_score_gemma":0.00008619592,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000008476745,"about_ca_topic_score_gemma":0.00001792818,"domain_scores_codex":[0.9994369,0.00001418308,0.0001670756,0.0001973162,0.00009020028,0.00009433885],"domain_scores_gemma":[0.9993873,0.0002521418,0.00008782811,0.0000976338,0.0001491712,0.00002593524],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","study_design_scores_codex":[0.00002817092,0.00001219172,7.819619e-7,0.0001723367,0.0000821318,0.000002205222,0.0000260711,0.0001518777,0.05979735,0.000002279298,0.9384154,0.001309177],"study_design_scores_gemma":[0.0002792031,0.0003358827,0.0002140226,0.002447786,0.00004446227,0.000009939928,0.0002116598,0.01726301,0.3311898,0.0001663357,0.6476052,0.0002327776],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"dataset","genre_gemma":"dataset","genre_scores_codex":[0.0006921382,0.0002449001,0.00006110244,0.00002695246,0.00001598669,0.0001348197,0.9981038,0.000007852615,0.0007124264],"genre_scores_gemma":[0.1118117,0.000004795569,0.000326966,0.00003479833,0.0001236068,0.0004404939,0.8858106,0.000008845952,0.001438214],"genre_candidate":"dataset","genre_consensus":"dataset","teacher_disagreement_score":0.7265865,"threshold_uncertainty_score":0.2792872,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02329691758867557,"score_gpt":0.2763404268834548,"score_spread":0.2530435092947792,"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."}}