{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:YKO25LPQ2RTVXOCXJLIA4MKMDF","short_pith_number":"pith:YKO25LPQ","schema_version":"1.0","canonical_sha256":"c29daeadf0d4675bb8574ad00e314c196b9382eeeca6e266646048aefaa69568","source":{"kind":"arxiv","id":"2311.14387","version":4},"attestation_state":"computed","paper":{"title":"Achieving Margin Maximization Exponentially Fast via Progressive Norm Rescaling","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["math.OC"],"primary_cat":"cs.LG","authors_text":"Lei Wu, Mingze Wang, Zeping Min","submitted_at":"2023-11-24T10:07:10Z","abstract_excerpt":"In this work, we investigate the margin-maximization bias exhibited by gradient-based algorithms in classifying linearly separable data. We present an in-depth analysis of the specific properties of the velocity field associated with (normalized) gradients, focusing on their role in margin maximization. Inspired by this analysis, we propose a novel algorithm called Progressive Rescaling Gradient Descent (PRGD) and show that PRGD can maximize the margin at an {\\em exponential rate}. This stands in stark contrast to all existing algorithms, which maximize the margin at a slow {\\em polynomial rat"},"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":"2311.14387","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2023-11-24T10:07:10Z","cross_cats_sorted":["math.OC"],"title_canon_sha256":"5a71ef71075a819255c09983a85714b26c973cf1f6a8a322c193d158ccd90aae","abstract_canon_sha256":"71f0e0bef9baa44efbb1a2f3964fe97f85ed56f839ec8143669ba15897c338ca"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:54:00.478363Z","signature_b64":"sxe+3WrRKTrG81OYZEGr3M9pv+Vqx685SbnGwhNf/S7sAZpbvjVwBo/nn3+7yv2P2UHK7gzVc99oG8BKelJGAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c29daeadf0d4675bb8574ad00e314c196b9382eeeca6e266646048aefaa69568","last_reissued_at":"2026-07-05T09:54:00.477907Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:54:00.477907Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Achieving Margin Maximization Exponentially Fast via Progressive Norm Rescaling","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["math.OC"],"primary_cat":"cs.LG","authors_text":"Lei Wu, Mingze Wang, Zeping Min","submitted_at":"2023-11-24T10:07:10Z","abstract_excerpt":"In this work, we investigate the margin-maximization bias exhibited by gradient-based algorithms in classifying linearly separable data. We present an in-depth analysis of the specific properties of the velocity field associated with (normalized) gradients, focusing on their role in margin maximization. Inspired by this analysis, we propose a novel algorithm called Progressive Rescaling Gradient Descent (PRGD) and show that PRGD can maximize the margin at an {\\em exponential rate}. This stands in stark contrast to all existing algorithms, which maximize the margin at a slow {\\em polynomial rat"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2311.14387","kind":"arxiv","version":4},"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/2311.14387/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":"2311.14387","created_at":"2026-07-05T09:54:00.477962+00:00"},{"alias_kind":"arxiv_version","alias_value":"2311.14387v4","created_at":"2026-07-05T09:54:00.477962+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2311.14387","created_at":"2026-07-05T09:54:00.477962+00:00"},{"alias_kind":"pith_short_12","alias_value":"YKO25LPQ2RTV","created_at":"2026-07-05T09:54:00.477962+00:00"},{"alias_kind":"pith_short_16","alias_value":"YKO25LPQ2RTVXOCX","created_at":"2026-07-05T09:54:00.477962+00:00"},{"alias_kind":"pith_short_8","alias_value":"YKO25LPQ","created_at":"2026-07-05T09:54:00.477962+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2501.04697","citing_title":"Grokking at the Edge of Numerical Stability","ref_index":13,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/YKO25LPQ2RTVXOCXJLIA4MKMDF","json":"https://pith.science/pith/YKO25LPQ2RTVXOCXJLIA4MKMDF.json","graph_json":"https://pith.science/api/pith-number/YKO25LPQ2RTVXOCXJLIA4MKMDF/graph.json","events_json":"https://pith.science/api/pith-number/YKO25LPQ2RTVXOCXJLIA4MKMDF/events.json","paper":"https://pith.science/paper/YKO25LPQ"},"agent_actions":{"view_html":"https://pith.science/pith/YKO25LPQ2RTVXOCXJLIA4MKMDF","download_json":"https://pith.science/pith/YKO25LPQ2RTVXOCXJLIA4MKMDF.json","view_paper":"https://pith.science/paper/YKO25LPQ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2311.14387&json=true","fetch_graph":"https://pith.science/api/pith-number/YKO25LPQ2RTVXOCXJLIA4MKMDF/graph.json","fetch_events":"https://pith.science/api/pith-number/YKO25LPQ2RTVXOCXJLIA4MKMDF/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/YKO25LPQ2RTVXOCXJLIA4MKMDF/action/timestamp_anchor","attest_storage":"https://pith.science/pith/YKO25LPQ2RTVXOCXJLIA4MKMDF/action/storage_attestation","attest_author":"https://pith.science/pith/YKO25LPQ2RTVXOCXJLIA4MKMDF/action/author_attestation","sign_citation":"https://pith.science/pith/YKO25LPQ2RTVXOCXJLIA4MKMDF/action/citation_signature","submit_replication":"https://pith.science/pith/YKO25LPQ2RTVXOCXJLIA4MKMDF/action/replication_record"}},"created_at":"2026-07-05T09:54:00.477962+00:00","updated_at":"2026-07-05T09:54:00.477962+00:00"}