{"id":"W4414968374","doi":"10.48550/arxiv.2509.03378","title":"Understanding and Improving Shampoo and SOAP via Kullback-Leibler Minimization","year":2025,"lang":"en","type":"preprint","venue":"ArXiv.org","topic":"Speech and dialogue systems","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"RIKEN; University of Central Florida; Government of Canada; Canadian Institute for Advanced Research","keywords":"Shampoo; Overhead (engineering); Minification; Divergence (linguistics); SOAP","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":[0.005243245,0.002362479,0.001653698,0.0008533444,0.0009586867,0.002127403,0.003154248,0.003069153,0.006734775],"category_scores_gemma":[0.02263959,0.001084717,0.001264479,0.0006454951,0.002765219,0.004485737,0.005189578,0.005481307,0.003419128],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001173185,"about_ca_system_score_gemma":0.004235129,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003744109,"about_ca_topic_score_gemma":0.00533435,"domain_scores_codex":[0.9973532,0.0009471339,0.00017919,0.0004635763,0.0008659053,0.0001910244],"domain_scores_gemma":[0.9930117,0.003978083,0.0004010444,0.001253415,0.001044544,0.0003112151],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.000372166,0.0002207432,0.002057557,0.0006500606,0.000163693,0.0002435347,0.0002889518,0.5710494,0.0074308,0.1087005,0.02487903,0.2839435],"study_design_scores_gemma":[0.00004418018,0.00008364151,0.0001466762,0.000051432,0.0000138479,0.00006003212,0.00002976148,0.9396254,0.002237968,0.054214,0.003467458,0.00002567989],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.005615617,0.0005503077,0.9867572,0.0008789846,0.0001413842,0.00007798345,0.000104549,0.002610913,0.003263097],"genre_scores_gemma":[0.2237175,0.0009951144,0.7608334,0.001997967,0.0002903729,0.0005146887,0.0008716086,0.002468512,0.008310902],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.006734775,"threshold_uncertainty_score":0.02772927,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1053909708859754,"score_gpt":0.2626024331985078,"score_spread":0.1572114623125324,"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."}}