{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:PLIXWXL44QHNFRAIU2TIOZOONN","short_pith_number":"pith:PLIXWXL4","schema_version":"1.0","canonical_sha256":"7ad17b5d7ce40ed2c408a6a68765ce6b48cbf0e9ac7ae45e73e703ef37b6dae2","source":{"kind":"arxiv","id":"2602.11557","version":2},"attestation_state":"computed","paper":{"title":"The Implicit Bias of Steepest Descent with Mini-batch Stochastic Gradient","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Difan Zou, Jichu Li, Xuan Tang","submitted_at":"2026-02-12T04:25:38Z","abstract_excerpt":"A variety of widely used optimization methods like SignSGD and Muon can be interpreted as instances of steepest descent under different norm-induced geometries. In this work, we study the implicit bias of mini-batch stochastic steepest descent in multi-class classification, characterizing how batch size, momentum, and variance reduction shape the limiting max-margin behavior and convergence rates under general entry-wise and Schatten-$p$ norms. We show that, without momentum, worst-case convergence and successful classification can only be guaranteed with full-batch gradient. In contrast, mome"},"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":"2602.11557","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-02-12T04:25:38Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"d5454afa813be3d349299febe6dc2621711072329399e0c8b89c5f82ff9ef253","abstract_canon_sha256":"c4d07c52e8e84a13c1852e213a599da966c94621a0704157bc0866b5725bba17"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-19T16:11:20.769368Z","signature_b64":"cua/5AyszovclVUhkltNNOfsiFoA6ObVvbHmJD4K49WD/Mezc8t1Y5h4Uf3tMs4O9Kkm3V5v/DBS9y29GVelBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7ad17b5d7ce40ed2c408a6a68765ce6b48cbf0e9ac7ae45e73e703ef37b6dae2","last_reissued_at":"2026-06-19T16:11:20.768942Z","signature_status":"signed_v1","first_computed_at":"2026-06-19T16:11:20.768942Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"The Implicit Bias of Steepest Descent with Mini-batch Stochastic Gradient","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Difan Zou, Jichu Li, Xuan Tang","submitted_at":"2026-02-12T04:25:38Z","abstract_excerpt":"A variety of widely used optimization methods like SignSGD and Muon can be interpreted as instances of steepest descent under different norm-induced geometries. In this work, we study the implicit bias of mini-batch stochastic steepest descent in multi-class classification, characterizing how batch size, momentum, and variance reduction shape the limiting max-margin behavior and convergence rates under general entry-wise and Schatten-$p$ norms. We show that, without momentum, worst-case convergence and successful classification can only be guaranteed with full-batch gradient. In contrast, mome"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2602.11557","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/2602.11557/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":"2602.11557","created_at":"2026-06-19T16:11:20.768998+00:00"},{"alias_kind":"arxiv_version","alias_value":"2602.11557v2","created_at":"2026-06-19T16:11:20.768998+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2602.11557","created_at":"2026-06-19T16:11:20.768998+00:00"},{"alias_kind":"pith_short_12","alias_value":"PLIXWXL44QHN","created_at":"2026-06-19T16:11:20.768998+00:00"},{"alias_kind":"pith_short_16","alias_value":"PLIXWXL44QHNFRAI","created_at":"2026-06-19T16:11:20.768998+00:00"},{"alias_kind":"pith_short_8","alias_value":"PLIXWXL4","created_at":"2026-06-19T16:11:20.768998+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/PLIXWXL44QHNFRAIU2TIOZOONN","json":"https://pith.science/pith/PLIXWXL44QHNFRAIU2TIOZOONN.json","graph_json":"https://pith.science/api/pith-number/PLIXWXL44QHNFRAIU2TIOZOONN/graph.json","events_json":"https://pith.science/api/pith-number/PLIXWXL44QHNFRAIU2TIOZOONN/events.json","paper":"https://pith.science/paper/PLIXWXL4"},"agent_actions":{"view_html":"https://pith.science/pith/PLIXWXL44QHNFRAIU2TIOZOONN","download_json":"https://pith.science/pith/PLIXWXL44QHNFRAIU2TIOZOONN.json","view_paper":"https://pith.science/paper/PLIXWXL4","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2602.11557&json=true","fetch_graph":"https://pith.science/api/pith-number/PLIXWXL44QHNFRAIU2TIOZOONN/graph.json","fetch_events":"https://pith.science/api/pith-number/PLIXWXL44QHNFRAIU2TIOZOONN/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/PLIXWXL44QHNFRAIU2TIOZOONN/action/timestamp_anchor","attest_storage":"https://pith.science/pith/PLIXWXL44QHNFRAIU2TIOZOONN/action/storage_attestation","attest_author":"https://pith.science/pith/PLIXWXL44QHNFRAIU2TIOZOONN/action/author_attestation","sign_citation":"https://pith.science/pith/PLIXWXL44QHNFRAIU2TIOZOONN/action/citation_signature","submit_replication":"https://pith.science/pith/PLIXWXL44QHNFRAIU2TIOZOONN/action/replication_record"}},"created_at":"2026-06-19T16:11:20.768998+00:00","updated_at":"2026-06-19T16:11:20.768998+00:00"}