{"id":"W6920880728","doi":"10.6084/m9.figshare.28059550","title":"Additional file 1 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":[0.00001674831,0.00007499419,0.0001190588,0.00004131137,0.00001948439,0.00001103437,0.00008455506,0.00007080453,0.7315079],"category_scores_gemma":[0.001124971,0.00006850558,0.0001330232,0.0001121359,0.00001867656,0.000008808772,0.00004505177,0.00003078746,0.00005514791],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000007557494,"about_ca_system_score_gemma":0.00008614882,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000008382683,"about_ca_topic_score_gemma":0.00001780854,"domain_scores_codex":[0.9994369,0.00001419659,0.0001670022,0.0001974014,0.00009018215,0.00009434605],"domain_scores_gemma":[0.9993862,0.0002543672,0.00008781437,0.00009777873,0.0001478817,0.0000259374],"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":[0.00002787495,0.00001214098,7.77296e-7,0.0001686933,0.00008181849,0.000002169145,0.00002594902,0.0001511268,0.05937919,0.000002282856,0.9388635,0.00128444],"study_design_scores_gemma":[0.0002799199,0.0003418291,0.0002188157,0.002492899,0.00004559206,0.00001019601,0.0002161228,0.01751683,0.3302517,0.0001685591,0.648221,0.0002365453],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"dataset","genre_gemma":"dataset","genre_scores_codex":[0.000686839,0.0002435399,0.00006073373,0.0000270661,0.00001582978,0.0001347606,0.9981154,0.000007858952,0.0007079616],"genre_scores_gemma":[0.1106567,0.000004725399,0.0003305794,0.00003480148,0.0001221324,0.0004484587,0.8869838,0.000008840575,0.001409978],"genre_candidate":"dataset","genre_consensus":"dataset","teacher_disagreement_score":0.7314527,"threshold_uncertainty_score":0.2793576,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02372716648541075,"score_gpt":0.2767611067718642,"score_spread":0.2530339402864534,"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."}}