{"id":"W4379162524","doi":"10.22541/au.168571970.03060289/v1","title":"Parsing Millions of URLs per Second","year":2023,"lang":"en","type":"preprint","venue":"","topic":"Algorithms and Data Compression","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Université TÉLUQ","funders":"","keywords":"Computer science; Parsing; Node (physics); Programming language; Operating system; Physics","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":[],"consensus_categories":[],"category_scores_codex":[0.0001911202,0.0001442665,0.0002381061,0.0001382744,0.00006378057,0.0001290609,0.001225744,0.0001411762,0.0002596736],"category_scores_gemma":[0.00001443208,0.0001179563,0.0001065058,0.0001285298,0.00003023576,0.0001852912,0.004173909,0.0002698391,0.0001366631],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00001720447,"about_ca_system_score_gemma":0.0001085838,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001539563,"about_ca_topic_score_gemma":0.00002255396,"domain_scores_codex":[0.9987823,0.00003498451,0.0002740408,0.0004821444,0.000245397,0.000181145],"domain_scores_gemma":[0.9984745,0.00008441491,0.0001279357,0.00116422,0.00007897319,0.0000699719],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00001230037,0.0005233065,0.001168005,0.001391465,0.0003671416,0.0001159149,0.005152884,0.008241192,0.006523223,0.1088076,0.6663956,0.2013014],"study_design_scores_gemma":[0.0005508502,0.0001063605,0.01718731,0.001181884,0.00004257806,0.00002455845,0.0001465174,0.7364126,0.0108575,0.1063172,0.1257223,0.001450313],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.004117875,0.0001993371,0.9877999,0.0005739463,0.00172938,0.0001498285,0.00008334911,0.0002630452,0.005083288],"genre_scores_gemma":[0.08567898,0.0001577857,0.8977708,0.0002990153,0.0003029778,0.00003322394,0.0001462862,0.00003934985,0.01557158],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.7281714,"threshold_uncertainty_score":0.5202479,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05754635032976943,"score_gpt":0.2922918177338762,"score_spread":0.2347454674041068,"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."}}