{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:G6USKLAG6I5W5DC5IJGKTYRBYG","short_pith_number":"pith:G6USKLAG","schema_version":"1.0","canonical_sha256":"37a9252c06f23b6e8c5d424ca9e221c1ac206d246be79cc330dd1f776b045336","source":{"kind":"arxiv","id":"2404.09730","version":3},"attestation_state":"computed","paper":{"title":"Convergence Analysis of Probability Flow ODE for Score-based Generative Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.NA","math.CA","math.NA"],"primary_cat":"cs.LG","authors_text":"Daniel Zhengyu Huang, Jiaoyang Huang, Zhengjiang Lin","submitted_at":"2024-04-15T12:29:28Z","abstract_excerpt":"Score-based generative models have emerged as a powerful approach for sampling high-dimensional probability distributions. Despite their effectiveness, their theoretical underpinnings remain relatively underdeveloped. In this work, we study the convergence properties of deterministic samplers based on probability flow ODEs from both theoretical and numerical perspectives. Assuming access to $L^2$-accurate estimates of the score function, we prove the total variation between the target and the generated data distributions can be bounded above by $\\mathcal{O}(d^{3/4}\\delta^{1/2})$ in the continu"},"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":"2404.09730","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-04-15T12:29:28Z","cross_cats_sorted":["cs.NA","math.CA","math.NA"],"title_canon_sha256":"ebc8929841d1e70116bfd183b806d25bbd11462259e21b2a743e5f8341272270","abstract_canon_sha256":"4917b7fa8a1d0f09c6c38710755a9bd4e490ecea250484172191d8e6110cc5bf"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:51:21.381806Z","signature_b64":"NtiQ4nASS5+bmPPITHxmWYZ8F+uOYyNPJypHABoZN4p3N80j/MJ/TW3/j2ZJryYwJvPCux5BQdNRVVWe0TvzAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"37a9252c06f23b6e8c5d424ca9e221c1ac206d246be79cc330dd1f776b045336","last_reissued_at":"2026-07-05T10:51:21.381299Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:51:21.381299Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Convergence Analysis of Probability Flow ODE for Score-based Generative Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.NA","math.CA","math.NA"],"primary_cat":"cs.LG","authors_text":"Daniel Zhengyu Huang, Jiaoyang Huang, Zhengjiang Lin","submitted_at":"2024-04-15T12:29:28Z","abstract_excerpt":"Score-based generative models have emerged as a powerful approach for sampling high-dimensional probability distributions. Despite their effectiveness, their theoretical underpinnings remain relatively underdeveloped. In this work, we study the convergence properties of deterministic samplers based on probability flow ODEs from both theoretical and numerical perspectives. Assuming access to $L^2$-accurate estimates of the score function, we prove the total variation between the target and the generated data distributions can be bounded above by $\\mathcal{O}(d^{3/4}\\delta^{1/2})$ in the continu"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2404.09730","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/2404.09730/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":"2404.09730","created_at":"2026-07-05T10:51:21.381370+00:00"},{"alias_kind":"arxiv_version","alias_value":"2404.09730v3","created_at":"2026-07-05T10:51:21.381370+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2404.09730","created_at":"2026-07-05T10:51:21.381370+00:00"},{"alias_kind":"pith_short_12","alias_value":"G6USKLAG6I5W","created_at":"2026-07-05T10:51:21.381370+00:00"},{"alias_kind":"pith_short_16","alias_value":"G6USKLAG6I5W5DC5","created_at":"2026-07-05T10:51:21.381370+00:00"},{"alias_kind":"pith_short_8","alias_value":"G6USKLAG","created_at":"2026-07-05T10:51:21.381370+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.23627","citing_title":"Diffusion Models Adapt to Low-Dimensional Structure Under Flexible Coefficient Choices","ref_index":20,"is_internal_anchor":false},{"citing_arxiv_id":"2607.02137","citing_title":"ART for Diffusion Sampling: Continuous-Time Control and Actor-Critic Learning","ref_index":3,"is_internal_anchor":false},{"citing_arxiv_id":"2601.18681","citing_title":"ART for Diffusion Sampling: A Reinforcement Learning Approach to Timestep Schedule","ref_index":4,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/G6USKLAG6I5W5DC5IJGKTYRBYG","json":"https://pith.science/pith/G6USKLAG6I5W5DC5IJGKTYRBYG.json","graph_json":"https://pith.science/api/pith-number/G6USKLAG6I5W5DC5IJGKTYRBYG/graph.json","events_json":"https://pith.science/api/pith-number/G6USKLAG6I5W5DC5IJGKTYRBYG/events.json","paper":"https://pith.science/paper/G6USKLAG"},"agent_actions":{"view_html":"https://pith.science/pith/G6USKLAG6I5W5DC5IJGKTYRBYG","download_json":"https://pith.science/pith/G6USKLAG6I5W5DC5IJGKTYRBYG.json","view_paper":"https://pith.science/paper/G6USKLAG","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2404.09730&json=true","fetch_graph":"https://pith.science/api/pith-number/G6USKLAG6I5W5DC5IJGKTYRBYG/graph.json","fetch_events":"https://pith.science/api/pith-number/G6USKLAG6I5W5DC5IJGKTYRBYG/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/G6USKLAG6I5W5DC5IJGKTYRBYG/action/timestamp_anchor","attest_storage":"https://pith.science/pith/G6USKLAG6I5W5DC5IJGKTYRBYG/action/storage_attestation","attest_author":"https://pith.science/pith/G6USKLAG6I5W5DC5IJGKTYRBYG/action/author_attestation","sign_citation":"https://pith.science/pith/G6USKLAG6I5W5DC5IJGKTYRBYG/action/citation_signature","submit_replication":"https://pith.science/pith/G6USKLAG6I5W5DC5IJGKTYRBYG/action/replication_record"}},"created_at":"2026-07-05T10:51:21.381370+00:00","updated_at":"2026-07-05T10:51:21.381370+00:00"}