{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2019:G2BSPFFDXR2DZJUMEIZUGBE3CU","short_pith_number":"pith:G2BSPFFD","schema_version":"1.0","canonical_sha256":"36832794a3bc743ca68c223343049b1533369fab52a82c839a491609dd292e3e","source":{"kind":"arxiv","id":"1910.05874","version":2},"attestation_state":"computed","paper":{"title":"Effects of Depth, Width, and Initialization: A Convergence Analysis of Layer-wise Training for Deep Linear Neural Networks","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.NA","math.NA","stat.ML"],"primary_cat":"cs.LG","authors_text":"Yeonjong Shin","submitted_at":"2019-10-14T00:50:55Z","abstract_excerpt":"Deep neural networks have been used in various machine learning applications and achieved tremendous empirical successes. However, training deep neural networks is a challenging task. Many alternatives have been proposed in place of end-to-end back-propagation. Layer-wise training is one of them, which trains a single layer at a time, rather than trains the whole layers simultaneously. In this paper, we study a layer-wise training using a block coordinate gradient descent (BCGD) for deep linear networks. We establish a general convergence analysis of BCGD and found the optimal learning rate, w"},"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":"1910.05874","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2019-10-14T00:50:55Z","cross_cats_sorted":["cs.NA","math.NA","stat.ML"],"title_canon_sha256":"922dc83827836562e3cd3725bb9da3e52262916bf9220c6cd346da09d5bda923","abstract_canon_sha256":"4999a88f71e27625d43acd47b1d2adbc1a43f623baec614bcc06b52cbe564bd8"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T01:33:34.187276Z","signature_b64":"BktkQxbWbAzdnV3UJb0Ed7DWbB+VJuw+WZ6LuByolfNZvv1zizwE0ky44EL8VNmQeFkaAck+rAd2RW4hWOIACA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"36832794a3bc743ca68c223343049b1533369fab52a82c839a491609dd292e3e","last_reissued_at":"2026-07-05T01:33:34.186929Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T01:33:34.186929Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Effects of Depth, Width, and Initialization: A Convergence Analysis of Layer-wise Training for Deep Linear Neural Networks","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.NA","math.NA","stat.ML"],"primary_cat":"cs.LG","authors_text":"Yeonjong Shin","submitted_at":"2019-10-14T00:50:55Z","abstract_excerpt":"Deep neural networks have been used in various machine learning applications and achieved tremendous empirical successes. However, training deep neural networks is a challenging task. Many alternatives have been proposed in place of end-to-end back-propagation. Layer-wise training is one of them, which trains a single layer at a time, rather than trains the whole layers simultaneously. In this paper, we study a layer-wise training using a block coordinate gradient descent (BCGD) for deep linear networks. We establish a general convergence analysis of BCGD and found the optimal learning rate, w"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1910.05874","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/1910.05874/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":"1910.05874","created_at":"2026-07-05T01:33:34.186991+00:00"},{"alias_kind":"arxiv_version","alias_value":"1910.05874v2","created_at":"2026-07-05T01:33:34.186991+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1910.05874","created_at":"2026-07-05T01:33:34.186991+00:00"},{"alias_kind":"pith_short_12","alias_value":"G2BSPFFDXR2D","created_at":"2026-07-05T01:33:34.186991+00:00"},{"alias_kind":"pith_short_16","alias_value":"G2BSPFFDXR2DZJUM","created_at":"2026-07-05T01:33:34.186991+00:00"},{"alias_kind":"pith_short_8","alias_value":"G2BSPFFD","created_at":"2026-07-05T01:33:34.186991+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2506.06179","citing_title":"A Theoretical Study of (Hyper) Self-Attention through the Lens of Interactions: Representation, Training, Generalization","ref_index":41,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/G2BSPFFDXR2DZJUMEIZUGBE3CU","json":"https://pith.science/pith/G2BSPFFDXR2DZJUMEIZUGBE3CU.json","graph_json":"https://pith.science/api/pith-number/G2BSPFFDXR2DZJUMEIZUGBE3CU/graph.json","events_json":"https://pith.science/api/pith-number/G2BSPFFDXR2DZJUMEIZUGBE3CU/events.json","paper":"https://pith.science/paper/G2BSPFFD"},"agent_actions":{"view_html":"https://pith.science/pith/G2BSPFFDXR2DZJUMEIZUGBE3CU","download_json":"https://pith.science/pith/G2BSPFFDXR2DZJUMEIZUGBE3CU.json","view_paper":"https://pith.science/paper/G2BSPFFD","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1910.05874&json=true","fetch_graph":"https://pith.science/api/pith-number/G2BSPFFDXR2DZJUMEIZUGBE3CU/graph.json","fetch_events":"https://pith.science/api/pith-number/G2BSPFFDXR2DZJUMEIZUGBE3CU/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/G2BSPFFDXR2DZJUMEIZUGBE3CU/action/timestamp_anchor","attest_storage":"https://pith.science/pith/G2BSPFFDXR2DZJUMEIZUGBE3CU/action/storage_attestation","attest_author":"https://pith.science/pith/G2BSPFFDXR2DZJUMEIZUGBE3CU/action/author_attestation","sign_citation":"https://pith.science/pith/G2BSPFFDXR2DZJUMEIZUGBE3CU/action/citation_signature","submit_replication":"https://pith.science/pith/G2BSPFFDXR2DZJUMEIZUGBE3CU/action/replication_record"}},"created_at":"2026-07-05T01:33:34.186991+00:00","updated_at":"2026-07-05T01:33:34.186991+00:00"}