{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:IHBTPCZCKY3SBLH7UICBCJTADJ","short_pith_number":"pith:IHBTPCZC","schema_version":"1.0","canonical_sha256":"41c3378b22563720acffa2041126601a673abbf89a5c9fc62a630b2f3ea2ed80","source":{"kind":"arxiv","id":"2206.03299","version":2},"attestation_state":"computed","paper":{"title":"Generalization Error Bounds for Deep Neural Networks Trained by SGD","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Chao Ma, Mingze Wang","submitted_at":"2022-06-07T13:46:10Z","abstract_excerpt":"Generalization error bounds for deep neural networks trained by stochastic gradient descent (SGD) are derived by combining a dynamical control of an appropriate parameter norm and the Rademacher complexity estimate based on parameter norms. The bounds explicitly depend on the loss along the training trajectory, and work for a wide range of network architectures including multilayer perceptron (MLP) and convolutional neural networks (CNN). Compared with other algorithm-depending generalization estimates such as uniform stability-based bounds, our bounds do not require $L$-smoothness of the nonc"},"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":"2206.03299","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2022-06-07T13:46:10Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"1a4326cedc53231745c7d845f424702bbfddb3a12a8c8e1f942e6f50901c7eaa","abstract_canon_sha256":"f2b85631c67cdff2e82211cad2cf6b4109919c432f710e8a48c2174126a7b50b"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:14:35.103972Z","signature_b64":"ATL9ptuMzg3FKwvFKjWwAHG044jssQiEtkF5Gs7xKWGErsO2nPbil7z/+HObqaeLEUX+zSbi0/sv1T7uoOKUBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"41c3378b22563720acffa2041126601a673abbf89a5c9fc62a630b2f3ea2ed80","last_reissued_at":"2026-07-05T06:14:35.103547Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:14:35.103547Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Generalization Error Bounds for Deep Neural Networks Trained by SGD","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Chao Ma, Mingze Wang","submitted_at":"2022-06-07T13:46:10Z","abstract_excerpt":"Generalization error bounds for deep neural networks trained by stochastic gradient descent (SGD) are derived by combining a dynamical control of an appropriate parameter norm and the Rademacher complexity estimate based on parameter norms. The bounds explicitly depend on the loss along the training trajectory, and work for a wide range of network architectures including multilayer perceptron (MLP) and convolutional neural networks (CNN). Compared with other algorithm-depending generalization estimates such as uniform stability-based bounds, our bounds do not require $L$-smoothness of the nonc"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2206.03299","kind":"arxiv","version":2},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2206.03299/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"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":"2206.03299","created_at":"2026-07-05T06:14:35.103605+00:00"},{"alias_kind":"arxiv_version","alias_value":"2206.03299v2","created_at":"2026-07-05T06:14:35.103605+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2206.03299","created_at":"2026-07-05T06:14:35.103605+00:00"},{"alias_kind":"pith_short_12","alias_value":"IHBTPCZCKY3S","created_at":"2026-07-05T06:14:35.103605+00:00"},{"alias_kind":"pith_short_16","alias_value":"IHBTPCZCKY3SBLH7","created_at":"2026-07-05T06:14:35.103605+00:00"},{"alias_kind":"pith_short_8","alias_value":"IHBTPCZC","created_at":"2026-07-05T06:14:35.103605+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2503.02129","citing_title":"Path Regularization: A Near-Complete and Optimal Nonasymptotic Generalization Theory for Multilayer Neural Networks and Double Descent Phenomenon","ref_index":13,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/IHBTPCZCKY3SBLH7UICBCJTADJ","json":"https://pith.science/pith/IHBTPCZCKY3SBLH7UICBCJTADJ.json","graph_json":"https://pith.science/api/pith-number/IHBTPCZCKY3SBLH7UICBCJTADJ/graph.json","events_json":"https://pith.science/api/pith-number/IHBTPCZCKY3SBLH7UICBCJTADJ/events.json","paper":"https://pith.science/paper/IHBTPCZC"},"agent_actions":{"view_html":"https://pith.science/pith/IHBTPCZCKY3SBLH7UICBCJTADJ","download_json":"https://pith.science/pith/IHBTPCZCKY3SBLH7UICBCJTADJ.json","view_paper":"https://pith.science/paper/IHBTPCZC","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2206.03299&json=true","fetch_graph":"https://pith.science/api/pith-number/IHBTPCZCKY3SBLH7UICBCJTADJ/graph.json","fetch_events":"https://pith.science/api/pith-number/IHBTPCZCKY3SBLH7UICBCJTADJ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/IHBTPCZCKY3SBLH7UICBCJTADJ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/IHBTPCZCKY3SBLH7UICBCJTADJ/action/storage_attestation","attest_author":"https://pith.science/pith/IHBTPCZCKY3SBLH7UICBCJTADJ/action/author_attestation","sign_citation":"https://pith.science/pith/IHBTPCZCKY3SBLH7UICBCJTADJ/action/citation_signature","submit_replication":"https://pith.science/pith/IHBTPCZCKY3SBLH7UICBCJTADJ/action/replication_record"}},"created_at":"2026-07-05T06:14:35.103605+00:00","updated_at":"2026-07-05T06:14:35.103605+00:00"}