{"id":"W6920848929","doi":"10.6084/m9.figshare.28059553","title":"Additional file 2 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":[0.00001665871,0.00007497082,0.0001190873,0.00004128486,0.00001948578,0.00001105784,0.00008450895,0.00007080493,0.7278312],"category_scores_gemma":[0.001111679,0.00006848363,0.000132998,0.0001120505,0.00001867631,0.000008814686,0.00004504557,0.00003078051,0.00005457136],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000007559615,"about_ca_system_score_gemma":0.00008577316,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000008465448,"about_ca_topic_score_gemma":0.00001791182,"domain_scores_codex":[0.999437,0.00001419187,0.0001670755,0.0001972564,0.00009016001,0.00009431196],"domain_scores_gemma":[0.9993901,0.0002508452,0.00008793498,0.000097595,0.0001475599,0.00002593374],"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":[0.0000278248,0.00001218025,7.558133e-7,0.0001707962,0.00008211012,0.000002235218,0.00002606281,0.0001523782,0.05767651,0.000002330469,0.9405226,0.001324258],"study_design_scores_gemma":[0.0002770961,0.0003361806,0.0002117694,0.002421684,0.00004467223,0.00001004768,0.000210391,0.0171122,0.3229136,0.0001670394,0.6560631,0.0002321199],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"dataset","genre_gemma":"dataset","genre_scores_codex":[0.0006825796,0.0002474516,0.00005947952,0.00002693992,0.00001592664,0.0001345801,0.9981011,0.000007860269,0.0007240387],"genre_scores_gemma":[0.1091917,0.000004852292,0.0003237795,0.00003483737,0.0001240256,0.0004404651,0.8884556,0.00000883709,0.001415914],"genre_candidate":"dataset","genre_consensus":"dataset","teacher_disagreement_score":0.7277767,"threshold_uncertainty_score":0.2792681,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02289172318688318,"score_gpt":0.2761819589774402,"score_spread":0.253290235790557,"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."}}