{"id":"W7067332926","doi":"","title":"The matrix Dyson equation for machine learning: Correlated linearizations and the test error in random features regression","year":2024,"lang":"en","type":"dissertation","venue":"eScholarship@McGill (McGill)","topic":"Advanced Electron Microscopy Techniques and Applications","field":"Biochemistry, Genetics and Molecular Biology","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University","funders":"","keywords":"Regression; Regression analysis; Matrix (chemical analysis); Linear regression; Feature (linguistics); Test (biology); Polynomial regression","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":true,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["sts"],"consensus_categories":[],"category_scores_codex":[0.00070087,0.0003729973,0.0002958499,0.00009353962,0.001336683,0.0001076018,0.0003061164,0.0005521043,0.000004402237],"category_scores_gemma":[0.001600489,0.0002410757,0.0001696246,0.0003005315,0.00009830132,0.00001479272,0.00008627328,0.001026756,0.000005495108],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00008066824,"about_ca_system_score_gemma":0.00005041556,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00006959077,"about_ca_topic_score_gemma":0.0009534812,"domain_scores_codex":[0.9982367,0.0001681296,0.0004728052,0.0006220924,0.0001674374,0.0003327928],"domain_scores_gemma":[0.9985678,0.0004626453,0.0003269578,0.0003917839,0.0001903384,0.00006046832],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.002474884,0.0001121032,0.00002411883,0.0001880531,0.0001323195,0.000002168134,0.00002554176,0.0006085447,0.9038647,0.04734981,0.0001118591,0.04510593],"study_design_scores_gemma":[0.005600928,0.0005217415,0.0001325912,0.0005238543,0.0004088387,0.00003399011,0.000339443,0.005810109,0.5329821,0.03529749,0.4173886,0.0009603166],"study_design_candidate":"bench_or_experimental","study_design_consensus":"bench_or_experimental","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.8918505,0.06967923,0.00116123,0.001951554,0.001889374,0.01681853,0.003745719,0.0007897147,0.01211421],"genre_scores_gemma":[0.9317484,0.009364714,0.00223324,0.0001192745,0.00009106065,0.001726484,0.008283486,0.0002049494,0.04622839],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.4172768,"threshold_uncertainty_score":0.9999635,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.00774967661722233,"score_gpt":0.3035001571535692,"score_spread":0.2957504805363468,"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."}}