{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2020:2KB2BC3GIS3HY22RWS5ZJ65LAK","short_pith_number":"pith:2KB2BC3G","schema_version":"1.0","canonical_sha256":"d283a08b6644b67c6b51b4bb94fbab02997f05706b560f93d824a1a5447f39d7","source":{"kind":"arxiv","id":"2004.08657","version":1},"attestation_state":"computed","paper":{"title":"On Tight Convergence Rates of Without-replacement SGD","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"math.OC","authors_text":"Kwangjun Ahn, Suvrit Sra","submitted_at":"2020-04-18T16:14:11Z","abstract_excerpt":"For solving finite-sum optimization problems, SGD without replacement sampling is empirically shown to outperform SGD. Denoting by $n$ the number of components in the cost and $K$ the number of epochs of the algorithm , several recent works have shown convergence rates of without-replacement SGD that have better dependency on $n$ and $K$ than the baseline rate of $O(1/(nK))$ for SGD. However, there are two main limitations shared among those works: the rates have extra poly-logarithmic factors on $nK$, and denoting by $\\kappa$ the condition number of the problem, the rates hold after $\\kappa^c"},"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":"2004.08657","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"math.OC","submitted_at":"2020-04-18T16:14:11Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"091209df78f6b810ecfee5f4747bbd720616094a86bc77d755d48b8f6ac68b14","abstract_canon_sha256":"c59bd9b627a881bac3bb1b6eec430a3c5984e98b1fb0e5c365a3fa7c63b051d4"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T00:56:08.199321Z","signature_b64":"khcadyEXk6FWQDwdQdFeg6C0gN8ULOYJ0IKXoFn5ikafpcMrHd+BPwf3SLZ7vQhEBDuru9/hnRyyAdvcb3CRCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d283a08b6644b67c6b51b4bb94fbab02997f05706b560f93d824a1a5447f39d7","last_reissued_at":"2026-07-05T00:56:08.198867Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T00:56:08.198867Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"On Tight Convergence Rates of Without-replacement SGD","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"math.OC","authors_text":"Kwangjun Ahn, Suvrit Sra","submitted_at":"2020-04-18T16:14:11Z","abstract_excerpt":"For solving finite-sum optimization problems, SGD without replacement sampling is empirically shown to outperform SGD. Denoting by $n$ the number of components in the cost and $K$ the number of epochs of the algorithm , several recent works have shown convergence rates of without-replacement SGD that have better dependency on $n$ and $K$ than the baseline rate of $O(1/(nK))$ for SGD. However, there are two main limitations shared among those works: the rates have extra poly-logarithmic factors on $nK$, and denoting by $\\kappa$ the condition number of the problem, the rates hold after $\\kappa^c"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2004.08657","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/2004.08657/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":"2004.08657","created_at":"2026-07-05T00:56:08.198927+00:00"},{"alias_kind":"arxiv_version","alias_value":"2004.08657v1","created_at":"2026-07-05T00:56:08.198927+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2004.08657","created_at":"2026-07-05T00:56:08.198927+00:00"},{"alias_kind":"pith_short_12","alias_value":"2KB2BC3GIS3H","created_at":"2026-07-05T00:56:08.198927+00:00"},{"alias_kind":"pith_short_16","alias_value":"2KB2BC3GIS3HY22R","created_at":"2026-07-05T00:56:08.198927+00:00"},{"alias_kind":"pith_short_8","alias_value":"2KB2BC3G","created_at":"2026-07-05T00:56:08.198927+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.30930","citing_title":"SGD at the Edge of Stability: Stochastic Stabilization with Large Learning Rates","ref_index":66,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/2KB2BC3GIS3HY22RWS5ZJ65LAK","json":"https://pith.science/pith/2KB2BC3GIS3HY22RWS5ZJ65LAK.json","graph_json":"https://pith.science/api/pith-number/2KB2BC3GIS3HY22RWS5ZJ65LAK/graph.json","events_json":"https://pith.science/api/pith-number/2KB2BC3GIS3HY22RWS5ZJ65LAK/events.json","paper":"https://pith.science/paper/2KB2BC3G"},"agent_actions":{"view_html":"https://pith.science/pith/2KB2BC3GIS3HY22RWS5ZJ65LAK","download_json":"https://pith.science/pith/2KB2BC3GIS3HY22RWS5ZJ65LAK.json","view_paper":"https://pith.science/paper/2KB2BC3G","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2004.08657&json=true","fetch_graph":"https://pith.science/api/pith-number/2KB2BC3GIS3HY22RWS5ZJ65LAK/graph.json","fetch_events":"https://pith.science/api/pith-number/2KB2BC3GIS3HY22RWS5ZJ65LAK/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/2KB2BC3GIS3HY22RWS5ZJ65LAK/action/timestamp_anchor","attest_storage":"https://pith.science/pith/2KB2BC3GIS3HY22RWS5ZJ65LAK/action/storage_attestation","attest_author":"https://pith.science/pith/2KB2BC3GIS3HY22RWS5ZJ65LAK/action/author_attestation","sign_citation":"https://pith.science/pith/2KB2BC3GIS3HY22RWS5ZJ65LAK/action/citation_signature","submit_replication":"https://pith.science/pith/2KB2BC3GIS3HY22RWS5ZJ65LAK/action/replication_record"}},"created_at":"2026-07-05T00:56:08.198927+00:00","updated_at":"2026-07-05T00:56:08.198927+00:00"}