{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2019:4XGLYLUHH7KHWLOUBHQMFJVRJ3","short_pith_number":"pith:4XGLYLUH","schema_version":"1.0","canonical_sha256":"e5ccbc2e873fd47b2dd409e0c2a6b14ed19c36c0b8d9f6dabc70df386c761155","source":{"kind":"arxiv","id":"1905.10899","version":3},"attestation_state":"computed","paper":{"title":"ODE Analysis of Stochastic Gradient Methods with Optimism and Anchoring for Minimax Problems","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Ernest K. Ryu, Kun Yuan, Wotao Yin","submitted_at":"2019-05-26T23:05:13Z","abstract_excerpt":"Despite remarkable empirical success, the training dynamics of generative adversarial networks (GAN), which involves solving a minimax game using stochastic gradients, is still poorly understood. In this work, we analyze last-iterate convergence of simultaneous gradient descent (simGD) and its variants under the assumption of convex-concavity, guided by a continuous-time analysis with differential equations. First, we show that simGD, as is, converges with stochastic sub-gradients under strict convexity in the primal variable. Second, we generalize optimistic simGD to accommodate an optimism r"},"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":"1905.10899","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2019-05-26T23:05:13Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"9bece6e0e524fd82fe2b331bf244e13e18a319459f446b9ad6cf8da58ff240fa","abstract_canon_sha256":"2a08a244b50a49f4d3555d29b82d3301771dc02647da417de2186735ba549dc5"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T01:41:48.200953Z","signature_b64":"MSJPhK246KutCvdGo0fcvbg3P7oNIIDbUt5beo48VvJ6kXDP9k3d/V8owwpX9sM1kEbcRUgcZ3v43f2/+8V/CA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e5ccbc2e873fd47b2dd409e0c2a6b14ed19c36c0b8d9f6dabc70df386c761155","last_reissued_at":"2026-07-05T01:41:48.200475Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T01:41:48.200475Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"ODE Analysis of Stochastic Gradient Methods with Optimism and Anchoring for Minimax Problems","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Ernest K. Ryu, Kun Yuan, Wotao Yin","submitted_at":"2019-05-26T23:05:13Z","abstract_excerpt":"Despite remarkable empirical success, the training dynamics of generative adversarial networks (GAN), which involves solving a minimax game using stochastic gradients, is still poorly understood. In this work, we analyze last-iterate convergence of simultaneous gradient descent (simGD) and its variants under the assumption of convex-concavity, guided by a continuous-time analysis with differential equations. First, we show that simGD, as is, converges with stochastic sub-gradients under strict convexity in the primal variable. Second, we generalize optimistic simGD to accommodate an optimism r"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1905.10899","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/1905.10899/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":"1905.10899","created_at":"2026-07-05T01:41:48.200530+00:00"},{"alias_kind":"arxiv_version","alias_value":"1905.10899v3","created_at":"2026-07-05T01:41:48.200530+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1905.10899","created_at":"2026-07-05T01:41:48.200530+00:00"},{"alias_kind":"pith_short_12","alias_value":"4XGLYLUHH7KH","created_at":"2026-07-05T01:41:48.200530+00:00"},{"alias_kind":"pith_short_16","alias_value":"4XGLYLUHH7KHWLOU","created_at":"2026-07-05T01:41:48.200530+00:00"},{"alias_kind":"pith_short_8","alias_value":"4XGLYLUH","created_at":"2026-07-05T01:41:48.200530+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":4,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.22763","citing_title":"Advancing Mathematics Research with AI-Driven Formal Proof Search","ref_index":52,"is_internal_anchor":false},{"citing_arxiv_id":"2605.22763","citing_title":"Advancing Mathematics Research with AI-Driven Formal Proof Search","ref_index":52,"is_internal_anchor":false},{"citing_arxiv_id":"2604.12235","citing_title":"Last-Iterate Convergence of Anchored Gradient Descent","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2604.16972","citing_title":"MCPO: Mastery-Consolidated Policy Optimization for Large Reasoning Models","ref_index":14,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/4XGLYLUHH7KHWLOUBHQMFJVRJ3","json":"https://pith.science/pith/4XGLYLUHH7KHWLOUBHQMFJVRJ3.json","graph_json":"https://pith.science/api/pith-number/4XGLYLUHH7KHWLOUBHQMFJVRJ3/graph.json","events_json":"https://pith.science/api/pith-number/4XGLYLUHH7KHWLOUBHQMFJVRJ3/events.json","paper":"https://pith.science/paper/4XGLYLUH"},"agent_actions":{"view_html":"https://pith.science/pith/4XGLYLUHH7KHWLOUBHQMFJVRJ3","download_json":"https://pith.science/pith/4XGLYLUHH7KHWLOUBHQMFJVRJ3.json","view_paper":"https://pith.science/paper/4XGLYLUH","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1905.10899&json=true","fetch_graph":"https://pith.science/api/pith-number/4XGLYLUHH7KHWLOUBHQMFJVRJ3/graph.json","fetch_events":"https://pith.science/api/pith-number/4XGLYLUHH7KHWLOUBHQMFJVRJ3/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/4XGLYLUHH7KHWLOUBHQMFJVRJ3/action/timestamp_anchor","attest_storage":"https://pith.science/pith/4XGLYLUHH7KHWLOUBHQMFJVRJ3/action/storage_attestation","attest_author":"https://pith.science/pith/4XGLYLUHH7KHWLOUBHQMFJVRJ3/action/author_attestation","sign_citation":"https://pith.science/pith/4XGLYLUHH7KHWLOUBHQMFJVRJ3/action/citation_signature","submit_replication":"https://pith.science/pith/4XGLYLUHH7KHWLOUBHQMFJVRJ3/action/replication_record"}},"created_at":"2026-07-05T01:41:48.200530+00:00","updated_at":"2026-07-05T01:41:48.200530+00:00"}