{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:77IRRG3UH4KEO3K6P47PDD6WR4","short_pith_number":"pith:77IRRG3U","schema_version":"1.0","canonical_sha256":"ffd1189b743f14476d5e7f3ef18fd68f1b564d5db5cb95795ebcac7c279bb78c","source":{"kind":"arxiv","id":"2403.07723","version":3},"attestation_state":"computed","paper":{"title":"On the Last-Iterate Convergence of Shuffling Gradient Methods","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["math.OC","stat.ML"],"primary_cat":"cs.LG","authors_text":"Zhengyuan Zhou, Zijian Liu","submitted_at":"2024-03-12T15:01:17Z","abstract_excerpt":"Shuffling gradient methods are widely used in modern machine learning tasks and include three popular implementations: Random Reshuffle (RR), Shuffle Once (SO), and Incremental Gradient (IG). Compared to the empirical success, the theoretical guarantee of shuffling gradient methods was not well-understood for a long time. Until recently, the convergence rates had just been established for the average iterate for convex functions and the last iterate for strongly convex problems (using squared distance as the metric). However, when using the function value gap as the convergence criterion, exis"},"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":"2403.07723","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-03-12T15:01:17Z","cross_cats_sorted":["math.OC","stat.ML"],"title_canon_sha256":"59e68a83026db00b9ca963b132fc46f90726eda55445884eddb076074e94d15c","abstract_canon_sha256":"c4c223b5678edb98711ef5dd7473a67aa393eb803e51e2eaf85ec1e9e4779cdd"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:28:07.807296Z","signature_b64":"irnsX9FTZ9mvDkTzvtSgF+y8Pdqlz6uM+uN5B5YlmIVNF21MDJJlaVVk1vV5Gu08H+GOxA/rN0JdAqoHpXrRCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ffd1189b743f14476d5e7f3ef18fd68f1b564d5db5cb95795ebcac7c279bb78c","last_reissued_at":"2026-07-05T08:28:07.806843Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:28:07.806843Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"On the Last-Iterate Convergence of Shuffling Gradient Methods","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["math.OC","stat.ML"],"primary_cat":"cs.LG","authors_text":"Zhengyuan Zhou, Zijian Liu","submitted_at":"2024-03-12T15:01:17Z","abstract_excerpt":"Shuffling gradient methods are widely used in modern machine learning tasks and include three popular implementations: Random Reshuffle (RR), Shuffle Once (SO), and Incremental Gradient (IG). Compared to the empirical success, the theoretical guarantee of shuffling gradient methods was not well-understood for a long time. Until recently, the convergence rates had just been established for the average iterate for convex functions and the last iterate for strongly convex problems (using squared distance as the metric). However, when using the function value gap as the convergence criterion, exis"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2403.07723","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/2403.07723/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":"2403.07723","created_at":"2026-07-05T08:28:07.806898+00:00"},{"alias_kind":"arxiv_version","alias_value":"2403.07723v3","created_at":"2026-07-05T08:28:07.806898+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2403.07723","created_at":"2026-07-05T08:28:07.806898+00:00"},{"alias_kind":"pith_short_12","alias_value":"77IRRG3UH4KE","created_at":"2026-07-05T08:28:07.806898+00:00"},{"alias_kind":"pith_short_16","alias_value":"77IRRG3UH4KEO3K6","created_at":"2026-07-05T08:28:07.806898+00:00"},{"alias_kind":"pith_short_8","alias_value":"77IRRG3U","created_at":"2026-07-05T08:28:07.806898+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2604.13870","citing_title":"Gradient Descent's Last Iterate is Often (slightly) Suboptimal","ref_index":16,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/77IRRG3UH4KEO3K6P47PDD6WR4","json":"https://pith.science/pith/77IRRG3UH4KEO3K6P47PDD6WR4.json","graph_json":"https://pith.science/api/pith-number/77IRRG3UH4KEO3K6P47PDD6WR4/graph.json","events_json":"https://pith.science/api/pith-number/77IRRG3UH4KEO3K6P47PDD6WR4/events.json","paper":"https://pith.science/paper/77IRRG3U"},"agent_actions":{"view_html":"https://pith.science/pith/77IRRG3UH4KEO3K6P47PDD6WR4","download_json":"https://pith.science/pith/77IRRG3UH4KEO3K6P47PDD6WR4.json","view_paper":"https://pith.science/paper/77IRRG3U","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2403.07723&json=true","fetch_graph":"https://pith.science/api/pith-number/77IRRG3UH4KEO3K6P47PDD6WR4/graph.json","fetch_events":"https://pith.science/api/pith-number/77IRRG3UH4KEO3K6P47PDD6WR4/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/77IRRG3UH4KEO3K6P47PDD6WR4/action/timestamp_anchor","attest_storage":"https://pith.science/pith/77IRRG3UH4KEO3K6P47PDD6WR4/action/storage_attestation","attest_author":"https://pith.science/pith/77IRRG3UH4KEO3K6P47PDD6WR4/action/author_attestation","sign_citation":"https://pith.science/pith/77IRRG3UH4KEO3K6P47PDD6WR4/action/citation_signature","submit_replication":"https://pith.science/pith/77IRRG3UH4KEO3K6P47PDD6WR4/action/replication_record"}},"created_at":"2026-07-05T08:28:07.806898+00:00","updated_at":"2026-07-05T08:28:07.806898+00:00"}