{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:DAQSY526SMJ6VCHMLVBRJURWLS","short_pith_number":"pith:DAQSY526","schema_version":"1.0","canonical_sha256":"18212c775e9313ea88ec5d4314d2365c9408f79fa935ae41a56f0d1e674caf23","source":{"kind":"arxiv","id":"2401.13115","version":3},"attestation_state":"computed","paper":{"title":"Contractive Diffusion Probabilistic Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Hanyang Zhao, Wenpin Tang","submitted_at":"2024-01-23T21:51:51Z","abstract_excerpt":"Diffusion probabilistic models (DPMs) have emerged as a promising technique in generative modeling. The success of DPMs relies on two ingredients: time reversal of diffusion processes and score matching. In view of possibly unguaranteed score matching, we propose a new criterion -- the contraction property of backward sampling in the design of DPMs, leading to a novel class of contractive DPMs (CDPMs). Our key insight is that, the contraction property can provably narrow score matching errors and discretization errors, thus our proposed CDPMs are robust to both sources of error. For practical "},"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":"2401.13115","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-01-23T21:51:51Z","cross_cats_sorted":[],"title_canon_sha256":"4f671b82a1cdd8ef8336368c85b851541c20d8696427f3919b1cc6259e3a8ab2","abstract_canon_sha256":"2fe387ac66447ec16f9b226919e36126d45e6b492c9319680ed9a48e6872da46"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:19:26.102906Z","signature_b64":"uet2/RmvUE21b6bSZMTSdgidVTaWaSj7Kb3s62QLJDS84CYZXqD8jhku8VzU0nXi6HwCMkzUA3rd9+oMqKOyBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"18212c775e9313ea88ec5d4314d2365c9408f79fa935ae41a56f0d1e674caf23","last_reissued_at":"2026-07-05T09:19:26.102422Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:19:26.102422Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Contractive Diffusion Probabilistic Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Hanyang Zhao, Wenpin Tang","submitted_at":"2024-01-23T21:51:51Z","abstract_excerpt":"Diffusion probabilistic models (DPMs) have emerged as a promising technique in generative modeling. The success of DPMs relies on two ingredients: time reversal of diffusion processes and score matching. In view of possibly unguaranteed score matching, we propose a new criterion -- the contraction property of backward sampling in the design of DPMs, leading to a novel class of contractive DPMs (CDPMs). Our key insight is that, the contraction property can provably narrow score matching errors and discretization errors, thus our proposed CDPMs are robust to both sources of error. For practical "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2401.13115","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/2401.13115/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":"2401.13115","created_at":"2026-07-05T09:19:26.102477+00:00"},{"alias_kind":"arxiv_version","alias_value":"2401.13115v3","created_at":"2026-07-05T09:19:26.102477+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2401.13115","created_at":"2026-07-05T09:19:26.102477+00:00"},{"alias_kind":"pith_short_12","alias_value":"DAQSY526SMJ6","created_at":"2026-07-05T09:19:26.102477+00:00"},{"alias_kind":"pith_short_16","alias_value":"DAQSY526SMJ6VCHM","created_at":"2026-07-05T09:19:26.102477+00:00"},{"alias_kind":"pith_short_8","alias_value":"DAQSY526","created_at":"2026-07-05T09:19:26.102477+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2607.08757","citing_title":"Score Accuracy Along the Forward Diffusion Does Not Certify Numerical Stability in Diffusion Sampling","ref_index":10,"is_internal_anchor":true},{"citing_arxiv_id":"2606.06007","citing_title":"Diffusion Models for Adaptive Sequential Data Generation","ref_index":65,"is_internal_anchor":false},{"citing_arxiv_id":"2512.18928","citing_title":"The Ensemble Schr{\\\"o}dinger Bridge filter for Nonlinear Data Assimilation","ref_index":36,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/DAQSY526SMJ6VCHMLVBRJURWLS","json":"https://pith.science/pith/DAQSY526SMJ6VCHMLVBRJURWLS.json","graph_json":"https://pith.science/api/pith-number/DAQSY526SMJ6VCHMLVBRJURWLS/graph.json","events_json":"https://pith.science/api/pith-number/DAQSY526SMJ6VCHMLVBRJURWLS/events.json","paper":"https://pith.science/paper/DAQSY526"},"agent_actions":{"view_html":"https://pith.science/pith/DAQSY526SMJ6VCHMLVBRJURWLS","download_json":"https://pith.science/pith/DAQSY526SMJ6VCHMLVBRJURWLS.json","view_paper":"https://pith.science/paper/DAQSY526","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2401.13115&json=true","fetch_graph":"https://pith.science/api/pith-number/DAQSY526SMJ6VCHMLVBRJURWLS/graph.json","fetch_events":"https://pith.science/api/pith-number/DAQSY526SMJ6VCHMLVBRJURWLS/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/DAQSY526SMJ6VCHMLVBRJURWLS/action/timestamp_anchor","attest_storage":"https://pith.science/pith/DAQSY526SMJ6VCHMLVBRJURWLS/action/storage_attestation","attest_author":"https://pith.science/pith/DAQSY526SMJ6VCHMLVBRJURWLS/action/author_attestation","sign_citation":"https://pith.science/pith/DAQSY526SMJ6VCHMLVBRJURWLS/action/citation_signature","submit_replication":"https://pith.science/pith/DAQSY526SMJ6VCHMLVBRJURWLS/action/replication_record"}},"created_at":"2026-07-05T09:19:26.102477+00:00","updated_at":"2026-07-05T09:19:26.102477+00:00"}