{"id":"W4399115775","doi":"10.48550/arxiv.2405.16563","title":"Higher-Order Transformer Derivative Estimates for Explicit Pathwise Learning Guarantees","year":2024,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Neural Networks and Applications","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"McMaster University; Natural Sciences and Engineering Research Council of Canada; Government of Canada; Canadian Institute for Advanced Research","keywords":"Generalization; Transformer; Path (computing); Mathematics; Computer science; Electrical engineering; Engineering; Mathematical analysis; Computer network; Voltage","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.005449004,0.002523597,0.001518416,0.001522848,0.0009007445,0.002400301,0.002957684,0.002160357,0.00886956],"category_scores_gemma":[0.04369726,0.0010702,0.001868769,0.001117857,0.003223785,0.01035065,0.005406982,0.007810954,0.001889939],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00395549,"about_ca_system_score_gemma":0.002334006,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00392195,"about_ca_topic_score_gemma":0.006128517,"domain_scores_codex":[0.9980076,0.0005612895,0.0001166563,0.0004984936,0.0005678929,0.0002481198],"domain_scores_gemma":[0.9774405,0.01724834,0.0009921896,0.002441244,0.001334085,0.0005437832],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.000250098,0.0001145126,0.002491277,0.0004246527,0.00008630421,0.0001994486,0.0003453387,0.5211628,0.00541204,0.3905974,0.006693914,0.07222233],"study_design_scores_gemma":[0.000006708016,0.00003010932,0.000233247,0.00004310774,0.00001409711,0.00004697735,0.00001549087,0.867288,0.001390558,0.1301061,0.0008117556,0.00001387557],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01495896,0.0008423434,0.9775724,0.0008006232,0.00005575789,0.00004297404,0.0002526442,0.0009343507,0.00454003],"genre_scores_gemma":[0.7060913,0.002747454,0.2748065,0.001219939,0.0003306066,0.0005001398,0.001339047,0.001909092,0.01105595],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.00886956,"threshold_uncertainty_score":0.02967161,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06907321891883815,"score_gpt":0.2131847142548939,"score_spread":0.1441114953360557,"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."}}