{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2019:B54UWSXA6N55FU2SE65CKZPUHT","short_pith_number":"pith:B54UWSXA","schema_version":"1.0","canonical_sha256":"0f794b4ae0f37bd2d35227ba2565f43cdfcaeab3c791720faa3c383189b789dd","source":{"kind":"arxiv","id":"1901.08276","version":1},"attestation_state":"computed","paper":{"title":"Traditional and Heavy-Tailed Self Regularization in Neural Network Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Charles H. Martin, Michael W. Mahoney","submitted_at":"2019-01-24T08:20:42Z","abstract_excerpt":"Random Matrix Theory (RMT) is applied to analyze the weight matrices of Deep Neural Networks (DNNs), including both production quality, pre-trained models such as AlexNet and Inception, and smaller models trained from scratch, such as LeNet5 and a miniature-AlexNet. Empirical and theoretical results clearly indicate that the empirical spectral density (ESD) of DNN layer matrices displays signatures of traditionally-regularized statistical models, even in the absence of exogenously specifying traditional forms of regularization, such as Dropout or Weight Norm constraints. Building on recent res"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"1901.08276","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2019-01-24T08:20:42Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"81e8226b24d614f04ad78b83629842294736040678978382d80d79080478db76","abstract_canon_sha256":"8494be1dc4b6bddd235c9bf7af83760bc2296a7b1095c96374dfc5e131120872"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:55:37.466802Z","signature_b64":"tET+visdhFSn1JDSNSBNtV83AxbhV9E1szGXYRE4XXPnT/eGAeOqvnGQJGTeqpvRfOR7o1gW0YsiaBVg943HAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0f794b4ae0f37bd2d35227ba2565f43cdfcaeab3c791720faa3c383189b789dd","last_reissued_at":"2026-05-17T23:55:37.466293Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:55:37.466293Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Traditional and Heavy-Tailed Self Regularization in Neural Network Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Charles H. Martin, Michael W. Mahoney","submitted_at":"2019-01-24T08:20:42Z","abstract_excerpt":"Random Matrix Theory (RMT) is applied to analyze the weight matrices of Deep Neural Networks (DNNs), including both production quality, pre-trained models such as AlexNet and Inception, and smaller models trained from scratch, such as LeNet5 and a miniature-AlexNet. Empirical and theoretical results clearly indicate that the empirical spectral density (ESD) of DNN layer matrices displays signatures of traditionally-regularized statistical models, even in the absence of exogenously specifying traditional forms of regularization, such as Dropout or Weight Norm constraints. Building on recent res"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1901.08276","kind":"arxiv","version":1},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"1901.08276","created_at":"2026-05-17T23:55:37.466394+00:00"},{"alias_kind":"arxiv_version","alias_value":"1901.08276v1","created_at":"2026-05-17T23:55:37.466394+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1901.08276","created_at":"2026-05-17T23:55:37.466394+00:00"},{"alias_kind":"pith_short_12","alias_value":"B54UWSXA6N55","created_at":"2026-05-18T12:33:12.712433+00:00"},{"alias_kind":"pith_short_16","alias_value":"B54UWSXA6N55FU2S","created_at":"2026-05-18T12:33:12.712433+00:00"},{"alias_kind":"pith_short_8","alias_value":"B54UWSXA","created_at":"2026-05-18T12:33:12.712433+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":8,"internal_anchor_count":8,"sample":[{"citing_arxiv_id":"2606.19268","citing_title":"Patnaik-Pearson intrinsic dimension for internal representations of neural networks","ref_index":23,"is_internal_anchor":true},{"citing_arxiv_id":"2606.19268","citing_title":"Patnaik-Pearson intrinsic dimension for internal representations of neural networks","ref_index":24,"is_internal_anchor":true},{"citing_arxiv_id":"2606.28486","citing_title":"Spectral phase transitions and trainability in neural network learning dynamics","ref_index":8,"is_internal_anchor":true},{"citing_arxiv_id":"2605.22297","citing_title":"One LR Doesn't Fit All: Heavy-Tail Guided Layerwise Learning Rates for LLMs","ref_index":10,"is_internal_anchor":true},{"citing_arxiv_id":"2412.00069","citing_title":"Condense, Don't Just Prune: Enhancing Efficiency and Performance in MoE Layer Pruning","ref_index":23,"is_internal_anchor":true},{"citing_arxiv_id":"2605.22297","citing_title":"One LR Doesn't Fit All: Heavy-Tail Guided Layerwise Learning Rates for LLMs","ref_index":11,"is_internal_anchor":true},{"citing_arxiv_id":"2605.18898","citing_title":"A Two-Parameter Weibull Framework for Diagnosing Transformer Weight Distributions","ref_index":3,"is_internal_anchor":true},{"citing_arxiv_id":"2604.13081","citing_title":"Selectivity and Shape in the Design of Forward-Forward Goodness Functions","ref_index":15,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/B54UWSXA6N55FU2SE65CKZPUHT","json":"https://pith.science/pith/B54UWSXA6N55FU2SE65CKZPUHT.json","graph_json":"https://pith.science/api/pith-number/B54UWSXA6N55FU2SE65CKZPUHT/graph.json","events_json":"https://pith.science/api/pith-number/B54UWSXA6N55FU2SE65CKZPUHT/events.json","paper":"https://pith.science/paper/B54UWSXA"},"agent_actions":{"view_html":"https://pith.science/pith/B54UWSXA6N55FU2SE65CKZPUHT","download_json":"https://pith.science/pith/B54UWSXA6N55FU2SE65CKZPUHT.json","view_paper":"https://pith.science/paper/B54UWSXA","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1901.08276&json=true","fetch_graph":"https://pith.science/api/pith-number/B54UWSXA6N55FU2SE65CKZPUHT/graph.json","fetch_events":"https://pith.science/api/pith-number/B54UWSXA6N55FU2SE65CKZPUHT/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/B54UWSXA6N55FU2SE65CKZPUHT/action/timestamp_anchor","attest_storage":"https://pith.science/pith/B54UWSXA6N55FU2SE65CKZPUHT/action/storage_attestation","attest_author":"https://pith.science/pith/B54UWSXA6N55FU2SE65CKZPUHT/action/author_attestation","sign_citation":"https://pith.science/pith/B54UWSXA6N55FU2SE65CKZPUHT/action/citation_signature","submit_replication":"https://pith.science/pith/B54UWSXA6N55FU2SE65CKZPUHT/action/replication_record"}},"created_at":"2026-05-17T23:55:37.466394+00:00","updated_at":"2026-05-17T23:55:37.466394+00:00"}