{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:LOHMWD45OJRV3F5NSDAEWB3RT5","short_pith_number":"pith:LOHMWD45","schema_version":"1.0","canonical_sha256":"5b8ecb0f9d72635d97ad90c04b07719f65d852fcc8bff5691f50541558beed5d","source":{"kind":"arxiv","id":"2504.21501","version":1},"attestation_state":"computed","paper":{"title":"Deep Learning Optimization Using Self-Adaptive Weighted Auxiliary Variables","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Michael K. Ng, Yaru Liu, Yiqi Gu","submitted_at":"2025-04-30T10:43:13Z","abstract_excerpt":"In this paper, we develop a new optimization framework for the least squares learning problem via fully connected neural networks or physics-informed neural networks. The gradient descent sometimes behaves inefficiently in deep learning because of the high non-convexity of loss functions and the vanishing gradient issue. Our idea is to introduce auxiliary variables to separate the layers of the deep neural networks and reformulate the loss functions for ease of optimization. We design the self-adaptive weights to preserve the consistency between the reformulated loss and the original mean squa"},"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":"2504.21501","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-04-30T10:43:13Z","cross_cats_sorted":[],"title_canon_sha256":"04680ea20821da39db28940783adfedbdb48a44bcea0920fecf4a7075c487f59","abstract_canon_sha256":"2239be2b58448d80834f8c2fbe78797444e5e50b377c6fdd01b3bbf34611060c"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:56:24.818982Z","signature_b64":"Rq9OjAz88FlUrUoZ/TYn2fP/DlySjMui9wmD9utfKUutBgwq7G0+14VhFWVJk4h0Ioc532PjFyRdnnLRMPYiBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5b8ecb0f9d72635d97ad90c04b07719f65d852fcc8bff5691f50541558beed5d","last_reissued_at":"2026-07-05T10:56:24.818471Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:56:24.818471Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Deep Learning Optimization Using Self-Adaptive Weighted Auxiliary Variables","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Michael K. Ng, Yaru Liu, Yiqi Gu","submitted_at":"2025-04-30T10:43:13Z","abstract_excerpt":"In this paper, we develop a new optimization framework for the least squares learning problem via fully connected neural networks or physics-informed neural networks. The gradient descent sometimes behaves inefficiently in deep learning because of the high non-convexity of loss functions and the vanishing gradient issue. Our idea is to introduce auxiliary variables to separate the layers of the deep neural networks and reformulate the loss functions for ease of optimization. We design the self-adaptive weights to preserve the consistency between the reformulated loss and the original mean squa"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2504.21501","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":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2504.21501/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":"2504.21501","created_at":"2026-07-05T10:56:24.818527+00:00"},{"alias_kind":"arxiv_version","alias_value":"2504.21501v1","created_at":"2026-07-05T10:56:24.818527+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2504.21501","created_at":"2026-07-05T10:56:24.818527+00:00"},{"alias_kind":"pith_short_12","alias_value":"LOHMWD45OJRV","created_at":"2026-07-05T10:56:24.818527+00:00"},{"alias_kind":"pith_short_16","alias_value":"LOHMWD45OJRV3F5N","created_at":"2026-07-05T10:56:24.818527+00:00"},{"alias_kind":"pith_short_8","alias_value":"LOHMWD45","created_at":"2026-07-05T10:56:24.818527+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2604.23225","citing_title":"A Layer Separation Optimization Framework for Cross-Entropy Training in Deep Learning","ref_index":2,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/LOHMWD45OJRV3F5NSDAEWB3RT5","json":"https://pith.science/pith/LOHMWD45OJRV3F5NSDAEWB3RT5.json","graph_json":"https://pith.science/api/pith-number/LOHMWD45OJRV3F5NSDAEWB3RT5/graph.json","events_json":"https://pith.science/api/pith-number/LOHMWD45OJRV3F5NSDAEWB3RT5/events.json","paper":"https://pith.science/paper/LOHMWD45"},"agent_actions":{"view_html":"https://pith.science/pith/LOHMWD45OJRV3F5NSDAEWB3RT5","download_json":"https://pith.science/pith/LOHMWD45OJRV3F5NSDAEWB3RT5.json","view_paper":"https://pith.science/paper/LOHMWD45","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2504.21501&json=true","fetch_graph":"https://pith.science/api/pith-number/LOHMWD45OJRV3F5NSDAEWB3RT5/graph.json","fetch_events":"https://pith.science/api/pith-number/LOHMWD45OJRV3F5NSDAEWB3RT5/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/LOHMWD45OJRV3F5NSDAEWB3RT5/action/timestamp_anchor","attest_storage":"https://pith.science/pith/LOHMWD45OJRV3F5NSDAEWB3RT5/action/storage_attestation","attest_author":"https://pith.science/pith/LOHMWD45OJRV3F5NSDAEWB3RT5/action/author_attestation","sign_citation":"https://pith.science/pith/LOHMWD45OJRV3F5NSDAEWB3RT5/action/citation_signature","submit_replication":"https://pith.science/pith/LOHMWD45OJRV3F5NSDAEWB3RT5/action/replication_record"}},"created_at":"2026-07-05T10:56:24.818527+00:00","updated_at":"2026-07-05T10:56:24.818527+00:00"}