{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:NJBBGLUE37XMBKJRM5SJXLCD2Q","short_pith_number":"pith:NJBBGLUE","schema_version":"1.0","canonical_sha256":"6a42132e84dfeec0a93167649bac43d42c1c4d0b2e969826dabe0078c772061e","source":{"kind":"arxiv","id":"2411.01803","version":2},"attestation_state":"computed","paper":{"title":"Gradient Methods with Online Scaling","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"math.OC","authors_text":"Madeleine Udell, Wenzhi Gao, Ya-Chi Chu, Yinyu Ye","submitted_at":"2024-11-04T05:04:18Z","abstract_excerpt":"We introduce a framework to accelerate the convergence of gradient-based methods with online learning. The framework learns to scale the gradient at each iteration through an online learning algorithm and provably accelerates gradient-based methods asymptotically. In contrast with previous literature, where convergence is established based on worst-case analysis, our framework provides a strong convergence guarantee with respect to the optimal scaling matrix for the iteration trajectory. For smooth strongly convex optimization, our results provide an $O(\\kappa^\\star \\log(1/\\varepsilon)$) compl"},"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":"2411.01803","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"math.OC","submitted_at":"2024-11-04T05:04:18Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"e61396710e8ba485199a561d1594764da300c9319623e7602b3754ad98227129","abstract_canon_sha256":"bcbf8a89eb0b88a378785f87e32c086e372bcf186a89d084224bcbd2ba01d8a9"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:31:31.236460Z","signature_b64":"wcYmbCIYsMNoFu/wekPIvF4esItx3XemFKraWwuPqrDQi74tBqyp+IKPttDuQ/mnLxqHD5sIsmcliS4De8VxAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6a42132e84dfeec0a93167649bac43d42c1c4d0b2e969826dabe0078c772061e","last_reissued_at":"2026-07-05T09:31:31.235942Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:31:31.235942Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Gradient Methods with Online Scaling","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"math.OC","authors_text":"Madeleine Udell, Wenzhi Gao, Ya-Chi Chu, Yinyu Ye","submitted_at":"2024-11-04T05:04:18Z","abstract_excerpt":"We introduce a framework to accelerate the convergence of gradient-based methods with online learning. The framework learns to scale the gradient at each iteration through an online learning algorithm and provably accelerates gradient-based methods asymptotically. In contrast with previous literature, where convergence is established based on worst-case analysis, our framework provides a strong convergence guarantee with respect to the optimal scaling matrix for the iteration trajectory. For smooth strongly convex optimization, our results provide an $O(\\kappa^\\star \\log(1/\\varepsilon)$) compl"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2411.01803","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/2411.01803/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":"2411.01803","created_at":"2026-07-05T09:31:31.235998+00:00"},{"alias_kind":"arxiv_version","alias_value":"2411.01803v2","created_at":"2026-07-05T09:31:31.235998+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2411.01803","created_at":"2026-07-05T09:31:31.235998+00:00"},{"alias_kind":"pith_short_12","alias_value":"NJBBGLUE37XM","created_at":"2026-07-05T09:31:31.235998+00:00"},{"alias_kind":"pith_short_16","alias_value":"NJBBGLUE37XMBKJR","created_at":"2026-07-05T09:31:31.235998+00:00"},{"alias_kind":"pith_short_8","alias_value":"NJBBGLUE","created_at":"2026-07-05T09:31:31.235998+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2510.10777","citing_title":"Preconditioned Norms: A Unified Framework for Steepest Descent, Quasi-Newton and Adaptive Methods","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2605.13473","citing_title":"OSDN: Improving Delta Rule with Provable Online Preconditioning in Linear Attention","ref_index":19,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/NJBBGLUE37XMBKJRM5SJXLCD2Q","json":"https://pith.science/pith/NJBBGLUE37XMBKJRM5SJXLCD2Q.json","graph_json":"https://pith.science/api/pith-number/NJBBGLUE37XMBKJRM5SJXLCD2Q/graph.json","events_json":"https://pith.science/api/pith-number/NJBBGLUE37XMBKJRM5SJXLCD2Q/events.json","paper":"https://pith.science/paper/NJBBGLUE"},"agent_actions":{"view_html":"https://pith.science/pith/NJBBGLUE37XMBKJRM5SJXLCD2Q","download_json":"https://pith.science/pith/NJBBGLUE37XMBKJRM5SJXLCD2Q.json","view_paper":"https://pith.science/paper/NJBBGLUE","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2411.01803&json=true","fetch_graph":"https://pith.science/api/pith-number/NJBBGLUE37XMBKJRM5SJXLCD2Q/graph.json","fetch_events":"https://pith.science/api/pith-number/NJBBGLUE37XMBKJRM5SJXLCD2Q/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/NJBBGLUE37XMBKJRM5SJXLCD2Q/action/timestamp_anchor","attest_storage":"https://pith.science/pith/NJBBGLUE37XMBKJRM5SJXLCD2Q/action/storage_attestation","attest_author":"https://pith.science/pith/NJBBGLUE37XMBKJRM5SJXLCD2Q/action/author_attestation","sign_citation":"https://pith.science/pith/NJBBGLUE37XMBKJRM5SJXLCD2Q/action/citation_signature","submit_replication":"https://pith.science/pith/NJBBGLUE37XMBKJRM5SJXLCD2Q/action/replication_record"}},"created_at":"2026-07-05T09:31:31.235998+00:00","updated_at":"2026-07-05T09:31:31.235998+00:00"}