{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2019:HL7SQKQ2ZLTHSEROQ4D3BTLMFH","short_pith_number":"pith:HL7SQKQ2","schema_version":"1.0","canonical_sha256":"3aff282a1acae679122e8707b0cd6c29c4c20611a42fb68d95b09d518cb7b0c6","source":{"kind":"arxiv","id":"1912.05475","version":3},"attestation_state":"computed","paper":{"title":"Mean-Field Neural ODEs via Relaxed Optimal Control","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["math.OC","stat.ML"],"primary_cat":"math.PR","authors_text":"David \\v{S}i\\v{s}ka, Jean-Fran\\c{c}ois Jabir, {\\L}ukasz Szpruch","submitted_at":"2019-12-11T17:13:06Z","abstract_excerpt":"We develop a framework for the analysis of deep neural networks and neural ODE models that are trained with stochastic gradient algorithms. We do that by identifying the connections between control theory, deep learning and theory of statistical sampling. We derive Pontryagin's optimality principle and study the corresponding gradient flow in the form of Mean-Field Langevin dynamics (MFLD) for solving relaxed data-driven control problems. Subsequently, we study uniform-in-time propagation of chaos of time-discretised MFLD. We derive explicit convergence rate in terms of the learning rate, the "},"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":"1912.05475","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"math.PR","submitted_at":"2019-12-11T17:13:06Z","cross_cats_sorted":["math.OC","stat.ML"],"title_canon_sha256":"5dca00e276ac8fbbf7932237cd7c89a0f580ba30cb76335c13289cbb0cab71fc","abstract_canon_sha256":"6679a2cd62cc78d2733cdaadc3b69ae63a96e032cab891b2f9b511e86de3f56e"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:23:56.111780Z","signature_b64":"1ajQV3jCDVddARBhkpPXZVO60LYUDlbPuVPTVPpu/adG+zw7JjigzG3WVda6UO3lI6IWUn0eJMdqGUaPwUy0Bg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3aff282a1acae679122e8707b0cd6c29c4c20611a42fb68d95b09d518cb7b0c6","last_reissued_at":"2026-07-05T02:23:56.111405Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:23:56.111405Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Mean-Field Neural ODEs via Relaxed Optimal Control","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["math.OC","stat.ML"],"primary_cat":"math.PR","authors_text":"David \\v{S}i\\v{s}ka, Jean-Fran\\c{c}ois Jabir, {\\L}ukasz Szpruch","submitted_at":"2019-12-11T17:13:06Z","abstract_excerpt":"We develop a framework for the analysis of deep neural networks and neural ODE models that are trained with stochastic gradient algorithms. We do that by identifying the connections between control theory, deep learning and theory of statistical sampling. We derive Pontryagin's optimality principle and study the corresponding gradient flow in the form of Mean-Field Langevin dynamics (MFLD) for solving relaxed data-driven control problems. Subsequently, we study uniform-in-time propagation of chaos of time-discretised MFLD. We derive explicit convergence rate in terms of the learning rate, the "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1912.05475","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/1912.05475/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":"1912.05475","created_at":"2026-07-05T02:23:56.111463+00:00"},{"alias_kind":"arxiv_version","alias_value":"1912.05475v3","created_at":"2026-07-05T02:23:56.111463+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1912.05475","created_at":"2026-07-05T02:23:56.111463+00:00"},{"alias_kind":"pith_short_12","alias_value":"HL7SQKQ2ZLTH","created_at":"2026-07-05T02:23:56.111463+00:00"},{"alias_kind":"pith_short_16","alias_value":"HL7SQKQ2ZLTHSERO","created_at":"2026-07-05T02:23:56.111463+00:00"},{"alias_kind":"pith_short_8","alias_value":"HL7SQKQ2","created_at":"2026-07-05T02:23:56.111463+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2503.15105","citing_title":"Control, Optimal Transport and Neural Differential Equations in Supervised Learning","ref_index":45,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/HL7SQKQ2ZLTHSEROQ4D3BTLMFH","json":"https://pith.science/pith/HL7SQKQ2ZLTHSEROQ4D3BTLMFH.json","graph_json":"https://pith.science/api/pith-number/HL7SQKQ2ZLTHSEROQ4D3BTLMFH/graph.json","events_json":"https://pith.science/api/pith-number/HL7SQKQ2ZLTHSEROQ4D3BTLMFH/events.json","paper":"https://pith.science/paper/HL7SQKQ2"},"agent_actions":{"view_html":"https://pith.science/pith/HL7SQKQ2ZLTHSEROQ4D3BTLMFH","download_json":"https://pith.science/pith/HL7SQKQ2ZLTHSEROQ4D3BTLMFH.json","view_paper":"https://pith.science/paper/HL7SQKQ2","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1912.05475&json=true","fetch_graph":"https://pith.science/api/pith-number/HL7SQKQ2ZLTHSEROQ4D3BTLMFH/graph.json","fetch_events":"https://pith.science/api/pith-number/HL7SQKQ2ZLTHSEROQ4D3BTLMFH/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/HL7SQKQ2ZLTHSEROQ4D3BTLMFH/action/timestamp_anchor","attest_storage":"https://pith.science/pith/HL7SQKQ2ZLTHSEROQ4D3BTLMFH/action/storage_attestation","attest_author":"https://pith.science/pith/HL7SQKQ2ZLTHSEROQ4D3BTLMFH/action/author_attestation","sign_citation":"https://pith.science/pith/HL7SQKQ2ZLTHSEROQ4D3BTLMFH/action/citation_signature","submit_replication":"https://pith.science/pith/HL7SQKQ2ZLTHSEROQ4D3BTLMFH/action/replication_record"}},"created_at":"2026-07-05T02:23:56.111463+00:00","updated_at":"2026-07-05T02:23:56.111463+00:00"}