{"id":"W2082347500","doi":"10.1098/rsta.2013.0138","title":"Large-scale detection of repetitions","year":2014,"lang":"en","type":"article","venue":"Philosophical Transactions of the Royal Society A Mathematical Physical and Engineering Sciences","topic":"Algorithms and Data Compression","field":"Computer Science","cited_by":3,"is_retracted":false,"has_abstract":true,"ca_institutions":"McMaster University","funders":"","keywords":"String (physics); Prefix; Computation; Suffix tree; Factorization; Suffix; Generalized suffix tree; Mathematics; Fraction (chemistry); Combinatorics; Trie; Computer science; Algorithm; Data structure","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001265163,0.0005779116,0.001231442,0.002056023,0.0008404155,0.001795453,0.001618669,0.001272742,0.002413465],"category_scores_gemma":[0.01898961,0.0006275031,0.0006279519,0.002168975,0.001248433,0.003900656,0.00222978,0.001340564,0.001631775],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006258833,"about_ca_system_score_gemma":0.0006412168,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0005301188,"about_ca_topic_score_gemma":0.0006160922,"domain_scores_codex":[0.9972434,0.0004775992,0.0001667363,0.0009277085,0.0009409165,0.0002436167],"domain_scores_gemma":[0.988704,0.005943061,0.001189921,0.002745157,0.00110578,0.0003119772],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.002221738,0.0002822692,0.03728803,0.0007362221,0.0002749626,0.002689698,0.001714488,0.08277328,0.208465,0.1255964,0.02147953,0.5164785],"study_design_scores_gemma":[0.00006761852,0.0002687004,0.009824918,0.0001088656,0.00006160261,0.002228066,0.0004761177,0.6943821,0.1002602,0.1773816,0.01483036,0.0001099188],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.4156517,0.001611919,0.5631048,0.001292707,0.0003313027,0.0001214491,0.001440951,0.005698544,0.01074667],"genre_scores_gemma":[0.8253076,0.0003842257,0.1684706,0.0002631812,0.0002214442,0.0001244378,0.00160073,0.0003789485,0.003248771],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.002413465,"threshold_uncertainty_score":0.008073807,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01044085314286231,"score_gpt":0.2177297521475103,"score_spread":0.207288899004648,"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."}}