{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:U2BHKMZEXIDNSC5PUPSXUL55R7","short_pith_number":"pith:U2BHKMZE","schema_version":"1.0","canonical_sha256":"a682753324ba06d90bafa3e57a2fbd8fef5aceb5ec6ec4c1d39a11fa3fe670be","source":{"kind":"arxiv","id":"2411.06476","version":3},"attestation_state":"computed","paper":{"title":"Eigen-componentwise convergence of SGD on quadratic programming","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.NA"],"primary_cat":"math.NA","authors_text":"Lehan Chen, Yuji Nakatsukasa","submitted_at":"2024-11-10T14:27:41Z","abstract_excerpt":"Stochastic gradient descent (SGD) is a workhorse algorithm for solving large-scale optimization problems in data science and machine learning. Understanding the convergence of SGD is hence of fundamental importance. In this work we examine the SGD convergence (with various step sizes) when applied to unconstrained convex quadratic programming (essentially least-squares (LS) problems), and in particular analyze the error components respect to the eigenvectors of the Hessian. The main message is that the convergence depends largely on the corresponding eigenvalues (singular values of the coeffic"},"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":"2411.06476","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"math.NA","submitted_at":"2024-11-10T14:27:41Z","cross_cats_sorted":["cs.NA"],"title_canon_sha256":"12531c45ef4963b98a75ba25d045b6f8234f484bd39909bde36d8ec2b743db46","abstract_canon_sha256":"f889605e8a280cd0f58a9cfa8cae736c2f46ccd478cbbd9a6a3003377fba92c7"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:46:53.658834Z","signature_b64":"e5YMqve0ZvSRp5+OgkkpSEQk4GDbdrWcYptwP0O4DLXVQN8s7Zz4zoYgrM944ATIvEpnQ6f05OhRMyC2bsZUDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a682753324ba06d90bafa3e57a2fbd8fef5aceb5ec6ec4c1d39a11fa3fe670be","last_reissued_at":"2026-07-05T09:46:53.657850Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:46:53.657850Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Eigen-componentwise convergence of SGD on quadratic programming","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.NA"],"primary_cat":"math.NA","authors_text":"Lehan Chen, Yuji Nakatsukasa","submitted_at":"2024-11-10T14:27:41Z","abstract_excerpt":"Stochastic gradient descent (SGD) is a workhorse algorithm for solving large-scale optimization problems in data science and machine learning. Understanding the convergence of SGD is hence of fundamental importance. In this work we examine the SGD convergence (with various step sizes) when applied to unconstrained convex quadratic programming (essentially least-squares (LS) problems), and in particular analyze the error components respect to the eigenvectors of the Hessian. The main message is that the convergence depends largely on the corresponding eigenvalues (singular values of the coeffic"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2411.06476","kind":"arxiv","version":3},"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/2411.06476/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":"2411.06476","created_at":"2026-07-05T09:46:53.658283+00:00"},{"alias_kind":"arxiv_version","alias_value":"2411.06476v3","created_at":"2026-07-05T09:46:53.658283+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2411.06476","created_at":"2026-07-05T09:46:53.658283+00:00"},{"alias_kind":"pith_short_12","alias_value":"U2BHKMZEXIDN","created_at":"2026-07-05T09:46:53.658283+00:00"},{"alias_kind":"pith_short_16","alias_value":"U2BHKMZEXIDNSC5P","created_at":"2026-07-05T09:46:53.658283+00:00"},{"alias_kind":"pith_short_8","alias_value":"U2BHKMZE","created_at":"2026-07-05T09:46:53.658283+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2501.15765","citing_title":"Hard edge asymptotics of correlation functions between singular values and eigenvalues","ref_index":13,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/U2BHKMZEXIDNSC5PUPSXUL55R7","json":"https://pith.science/pith/U2BHKMZEXIDNSC5PUPSXUL55R7.json","graph_json":"https://pith.science/api/pith-number/U2BHKMZEXIDNSC5PUPSXUL55R7/graph.json","events_json":"https://pith.science/api/pith-number/U2BHKMZEXIDNSC5PUPSXUL55R7/events.json","paper":"https://pith.science/paper/U2BHKMZE"},"agent_actions":{"view_html":"https://pith.science/pith/U2BHKMZEXIDNSC5PUPSXUL55R7","download_json":"https://pith.science/pith/U2BHKMZEXIDNSC5PUPSXUL55R7.json","view_paper":"https://pith.science/paper/U2BHKMZE","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2411.06476&json=true","fetch_graph":"https://pith.science/api/pith-number/U2BHKMZEXIDNSC5PUPSXUL55R7/graph.json","fetch_events":"https://pith.science/api/pith-number/U2BHKMZEXIDNSC5PUPSXUL55R7/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/U2BHKMZEXIDNSC5PUPSXUL55R7/action/timestamp_anchor","attest_storage":"https://pith.science/pith/U2BHKMZEXIDNSC5PUPSXUL55R7/action/storage_attestation","attest_author":"https://pith.science/pith/U2BHKMZEXIDNSC5PUPSXUL55R7/action/author_attestation","sign_citation":"https://pith.science/pith/U2BHKMZEXIDNSC5PUPSXUL55R7/action/citation_signature","submit_replication":"https://pith.science/pith/U2BHKMZEXIDNSC5PUPSXUL55R7/action/replication_record"}},"created_at":"2026-07-05T09:46:53.658283+00:00","updated_at":"2026-07-05T09:46:53.658283+00:00"}