{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:Q6GUUZX52AZZ4LVFKDV3IXMVYY","short_pith_number":"pith:Q6GUUZX5","schema_version":"1.0","canonical_sha256":"878d4a66fdd0339e2ea550ebb45d95c6373cbabf232229633b03e87fb931ee28","source":{"kind":"arxiv","id":"2405.15754","version":1},"attestation_state":"computed","paper":{"title":"Score-based generative models are provably robust: an uncertainty quantification perspective","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG","math.ST","stat.TH"],"primary_cat":"stat.ML","authors_text":"Benjamin J. Zhang, Markos A. Katsoulakis, Nikiforos Mimikos-Stamatopoulos","submitted_at":"2024-05-24T17:50:17Z","abstract_excerpt":"Through an uncertainty quantification (UQ) perspective, we show that score-based generative models (SGMs) are provably robust to the multiple sources of error in practical implementation. Our primary tool is the Wasserstein uncertainty propagation (WUP) theorem, a model-form UQ bound that describes how the $L^2$ error from learning the score function propagates to a Wasserstein-1 ($\\mathbf{d}_1$) ball around the true data distribution under the evolution of the Fokker-Planck equation. We show how errors due to (a) finite sample approximation, (b) early stopping, (c) score-matching objective ch"},"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":"2405.15754","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"stat.ML","submitted_at":"2024-05-24T17:50:17Z","cross_cats_sorted":["cs.LG","math.ST","stat.TH"],"title_canon_sha256":"1aa29ec627f07c37259763d5136043683cfe8c4c738d82d5d00d38c1c7dca8ff","abstract_canon_sha256":"4d3c9979378b44794fb2816123ab3360b61c5015701600f3ebf0e68a4b1b8341"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:22:53.622332Z","signature_b64":"Bf7XCTpTrJp3mSTnEs9n1YMtj/FCrujvitFkEIjIW95tqfuOHSbPyRRwJe2i9t3ZdfDCeBJ2+S1eNdCoPF2QCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"878d4a66fdd0339e2ea550ebb45d95c6373cbabf232229633b03e87fb931ee28","last_reissued_at":"2026-07-05T08:22:53.621820Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:22:53.621820Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Score-based generative models are provably robust: an uncertainty quantification perspective","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG","math.ST","stat.TH"],"primary_cat":"stat.ML","authors_text":"Benjamin J. Zhang, Markos A. Katsoulakis, Nikiforos Mimikos-Stamatopoulos","submitted_at":"2024-05-24T17:50:17Z","abstract_excerpt":"Through an uncertainty quantification (UQ) perspective, we show that score-based generative models (SGMs) are provably robust to the multiple sources of error in practical implementation. Our primary tool is the Wasserstein uncertainty propagation (WUP) theorem, a model-form UQ bound that describes how the $L^2$ error from learning the score function propagates to a Wasserstein-1 ($\\mathbf{d}_1$) ball around the true data distribution under the evolution of the Fokker-Planck equation. We show how errors due to (a) finite sample approximation, (b) early stopping, (c) score-matching objective ch"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2405.15754","kind":"arxiv","version":1},"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/2405.15754/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":"2405.15754","created_at":"2026-07-05T08:22:53.621881+00:00"},{"alias_kind":"arxiv_version","alias_value":"2405.15754v1","created_at":"2026-07-05T08:22:53.621881+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2405.15754","created_at":"2026-07-05T08:22:53.621881+00:00"},{"alias_kind":"pith_short_12","alias_value":"Q6GUUZX52AZZ","created_at":"2026-07-05T08:22:53.621881+00:00"},{"alias_kind":"pith_short_16","alias_value":"Q6GUUZX52AZZ4LVF","created_at":"2026-07-05T08:22:53.621881+00:00"},{"alias_kind":"pith_short_8","alias_value":"Q6GUUZX5","created_at":"2026-07-05T08:22:53.621881+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2607.02349","citing_title":"A PDE-Based Framework for Generative Modeling Beyond Classical Score-Based Diffusion","ref_index":24,"is_internal_anchor":false},{"citing_arxiv_id":"2410.01244","citing_title":"Robustness and Structure Preservation in Flow-Based Generative Models via Wasserstein Path-Space Divergences","ref_index":23,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/Q6GUUZX52AZZ4LVFKDV3IXMVYY","json":"https://pith.science/pith/Q6GUUZX52AZZ4LVFKDV3IXMVYY.json","graph_json":"https://pith.science/api/pith-number/Q6GUUZX52AZZ4LVFKDV3IXMVYY/graph.json","events_json":"https://pith.science/api/pith-number/Q6GUUZX52AZZ4LVFKDV3IXMVYY/events.json","paper":"https://pith.science/paper/Q6GUUZX5"},"agent_actions":{"view_html":"https://pith.science/pith/Q6GUUZX52AZZ4LVFKDV3IXMVYY","download_json":"https://pith.science/pith/Q6GUUZX52AZZ4LVFKDV3IXMVYY.json","view_paper":"https://pith.science/paper/Q6GUUZX5","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2405.15754&json=true","fetch_graph":"https://pith.science/api/pith-number/Q6GUUZX52AZZ4LVFKDV3IXMVYY/graph.json","fetch_events":"https://pith.science/api/pith-number/Q6GUUZX52AZZ4LVFKDV3IXMVYY/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/Q6GUUZX52AZZ4LVFKDV3IXMVYY/action/timestamp_anchor","attest_storage":"https://pith.science/pith/Q6GUUZX52AZZ4LVFKDV3IXMVYY/action/storage_attestation","attest_author":"https://pith.science/pith/Q6GUUZX52AZZ4LVFKDV3IXMVYY/action/author_attestation","sign_citation":"https://pith.science/pith/Q6GUUZX52AZZ4LVFKDV3IXMVYY/action/citation_signature","submit_replication":"https://pith.science/pith/Q6GUUZX52AZZ4LVFKDV3IXMVYY/action/replication_record"}},"created_at":"2026-07-05T08:22:53.621881+00:00","updated_at":"2026-07-05T08:22:53.621881+00:00"}