{"id":"W4414812742","doi":"10.48550/arxiv.2507.09029","title":"Model Parallelism With Subnetwork Data Parallelism","year":2025,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Distributed and Parallel Computing Systems","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Natural Sciences and Engineering Research Council of Canada; Fonds de recherche du Québec – Nature et technologies; Alliance de recherche numérique du Canada; Agence Nationale de la Recherche","keywords":"Subnetwork; Block (permutation group theory); Masking (illustration); Parallelism (grammar); Scale (ratio); Data parallelism","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":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow","open_science"],"consensus_categories":["open_science"],"category_scores_codex":[0.0005199102,0.0006202347,0.0006939755,0.0002377849,0.0002933234,0.0003920146,0.009508037,0.0004629812,0.000007496552],"category_scores_gemma":[0.00002139754,0.0006275743,0.0001763364,0.0008998517,0.0001261188,0.0005406125,0.008711648,0.0009626542,0.00006264164],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001498382,"about_ca_system_score_gemma":0.000835322,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0002811859,"about_ca_topic_score_gemma":0.0001075659,"domain_scores_codex":[0.9959187,0.0002409142,0.0003841031,0.002517498,0.0002184606,0.0007203194],"domain_scores_gemma":[0.99317,0.0001572934,0.0004102686,0.005758272,0.0002398078,0.0002643916],"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":[0.00003115095,0.00006764171,0.0005260374,0.00009760679,0.000164504,0.0002354642,0.0001265195,0.8298544,2.685615e-7,0.1580531,0.01060524,0.000238066],"study_design_scores_gemma":[0.0006042683,0.00003199597,0.000124358,0.0003665535,0.0000969862,0.000009327138,0.00001621661,0.9650615,7.885558e-7,0.0266934,0.006287054,0.0007075081],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.008566977,0.0002939957,0.9780395,0.0002786318,0.0008747519,0.000473907,0.0002323808,0.0006946312,0.01054525],"genre_scores_gemma":[0.9755876,0.0002021225,0.01401814,0.0002370443,0.0001871098,0.00000210654,0.0005197995,0.00002108962,0.009224974],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9670206,"threshold_uncertainty_score":0.9996176,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1481211918186507,"score_gpt":0.2098409971280553,"score_spread":0.06171980530940463,"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."}}