{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:TAP5RIDSYCFKL6DBMLMASSG45Z","short_pith_number":"pith:TAP5RIDS","schema_version":"1.0","canonical_sha256":"981fd8a072c08aa5f86162d80948dcee6c41e2101976a0866a6603a8bf391891","source":{"kind":"arxiv","id":"2102.09235","version":3},"attestation_state":"computed","paper":{"title":"Deep Residual Networks Learn the Geodesic Curve in the Wasserstein Space","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Kuo Gai, Shihua Zhang","submitted_at":"2021-02-18T09:37:49Z","abstract_excerpt":"Recent studies revealed the mathematical connection between deep neural networks (DNNs) and dynamic systems. However, the specific dynamics that DNNs, especially deep residual networks (ResNets), tend to learn during training remain insufficiently characterized. To this end, we model the forward propagation of deep residual networks using continuity equations, in which the measure is conserved and infinite curves in the measure space connect the input distribution to the output one of a ResNet. We find ResNets with $L_2$ regularization attempt to learn the geodesic curve in the Wasserstein spa"},"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":"2102.09235","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2021-02-18T09:37:49Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"ba6ed1322b0e5283ef156c049cabd16dc1684bf5e1e7472258a397d7a7aeae24","abstract_canon_sha256":"17862507dac2ebc93b2055944416b973e02eef9d111549aaba567b65ebb16819"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-29T01:15:01.537351Z","signature_b64":"INrJ2nLsaNk/dWkQ1D+OYDoRLXBS5QC5VecmXRcWWO0+N8GHbJCUcigYKmFdcx+FWSVIkw12hU6H9zR7jwEbDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"981fd8a072c08aa5f86162d80948dcee6c41e2101976a0866a6603a8bf391891","last_reissued_at":"2026-06-29T01:15:01.536932Z","signature_status":"signed_v1","first_computed_at":"2026-06-29T01:15:01.536932Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Deep Residual Networks Learn the Geodesic Curve in the Wasserstein Space","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Kuo Gai, Shihua Zhang","submitted_at":"2021-02-18T09:37:49Z","abstract_excerpt":"Recent studies revealed the mathematical connection between deep neural networks (DNNs) and dynamic systems. However, the specific dynamics that DNNs, especially deep residual networks (ResNets), tend to learn during training remain insufficiently characterized. To this end, we model the forward propagation of deep residual networks using continuity equations, in which the measure is conserved and infinite curves in the measure space connect the input distribution to the output one of a ResNet. We find ResNets with $L_2$ regularization attempt to learn the geodesic curve in the Wasserstein spa"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2102.09235","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/2102.09235/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":"2102.09235","created_at":"2026-06-29T01:15:01.536990+00:00"},{"alias_kind":"arxiv_version","alias_value":"2102.09235v3","created_at":"2026-06-29T01:15:01.536990+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2102.09235","created_at":"2026-06-29T01:15:01.536990+00:00"},{"alias_kind":"pith_short_12","alias_value":"TAP5RIDSYCFK","created_at":"2026-06-29T01:15:01.536990+00:00"},{"alias_kind":"pith_short_16","alias_value":"TAP5RIDSYCFKL6DB","created_at":"2026-06-29T01:15:01.536990+00:00"},{"alias_kind":"pith_short_8","alias_value":"TAP5RIDS","created_at":"2026-06-29T01:15:01.536990+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2606.30813","citing_title":"Gradient Smoothing: Coupling Layer-wise Updates for Improved Optimization","ref_index":6,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/TAP5RIDSYCFKL6DBMLMASSG45Z","json":"https://pith.science/pith/TAP5RIDSYCFKL6DBMLMASSG45Z.json","graph_json":"https://pith.science/api/pith-number/TAP5RIDSYCFKL6DBMLMASSG45Z/graph.json","events_json":"https://pith.science/api/pith-number/TAP5RIDSYCFKL6DBMLMASSG45Z/events.json","paper":"https://pith.science/paper/TAP5RIDS"},"agent_actions":{"view_html":"https://pith.science/pith/TAP5RIDSYCFKL6DBMLMASSG45Z","download_json":"https://pith.science/pith/TAP5RIDSYCFKL6DBMLMASSG45Z.json","view_paper":"https://pith.science/paper/TAP5RIDS","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2102.09235&json=true","fetch_graph":"https://pith.science/api/pith-number/TAP5RIDSYCFKL6DBMLMASSG45Z/graph.json","fetch_events":"https://pith.science/api/pith-number/TAP5RIDSYCFKL6DBMLMASSG45Z/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/TAP5RIDSYCFKL6DBMLMASSG45Z/action/timestamp_anchor","attest_storage":"https://pith.science/pith/TAP5RIDSYCFKL6DBMLMASSG45Z/action/storage_attestation","attest_author":"https://pith.science/pith/TAP5RIDSYCFKL6DBMLMASSG45Z/action/author_attestation","sign_citation":"https://pith.science/pith/TAP5RIDSYCFKL6DBMLMASSG45Z/action/citation_signature","submit_replication":"https://pith.science/pith/TAP5RIDSYCFKL6DBMLMASSG45Z/action/replication_record"}},"created_at":"2026-06-29T01:15:01.536990+00:00","updated_at":"2026-06-29T01:15:01.536990+00:00"}