{"id":"W3033781806","doi":"10.1007/s10208-022-09554-y","title":"Halting Time is Predictable for Large Models: A Universality Property and Average-Case Analysis","year":2022,"lang":"en","type":"article","venue":"Foundations of Computational Mathematics","topic":"Stochastic Gradient Optimization Techniques","field":"Computer Science","cited_by":15,"is_retracted":false,"has_abstract":false,"ca_institutions":"McGill University; Google (Canada)","funders":"","keywords":"Universality (dynamical systems); Mathematics; Case analysis; Property (philosophy); Probability distribution; Applied mathematics; Algorithm; Mathematical optimization; Statistics; Computer science; Artificial intelligence","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":[0.01235723,0.002569139,0.00612124,0.003778724,0.002671188,0.005905473,0.005601698,0.003568325,0.00793564],"category_scores_gemma":[0.07954174,0.00236285,0.004787073,0.002704891,0.009923915,0.01828285,0.006283417,0.009973447,0.0005198939],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.003465791,"about_ca_system_score_gemma":0.003381126,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003103635,"about_ca_topic_score_gemma":0.002646223,"domain_scores_codex":[0.9936433,0.001867565,0.0003935011,0.001836114,0.00105611,0.001203431],"domain_scores_gemma":[0.8846212,0.08765104,0.009662263,0.008194526,0.004655721,0.005215311],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0001567899,0.0001718343,0.003349331,0.000319205,0.0002819011,0.0006317695,0.0005546357,0.07180015,0.001621731,0.9129745,0.002315831,0.005822274],"study_design_scores_gemma":[0.00003009234,0.00004905953,0.0006708134,0.00004251145,0.00008950595,0.0002003518,0.00008192562,0.3871543,0.0005784601,0.6106921,0.0003556417,0.00005524134],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1723726,0.002014379,0.8058662,0.004126514,0.000198333,0.0001589412,0.0006277579,0.001148766,0.01348651],"genre_scores_gemma":[0.9683782,0.001537835,0.0227447,0.0006886,0.0006579697,0.000337913,0.0004452449,0.0005081713,0.004701286],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01235723,"threshold_uncertainty_score":0.06535208,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03203514073703707,"score_gpt":0.2679588309108773,"score_spread":0.2359236901738402,"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."}}