{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:NVCEEXBKVFWTWSMD6HDLIYYJHU","short_pith_number":"pith:NVCEEXBK","schema_version":"1.0","canonical_sha256":"6d44425c2aa96d3b4983f1c6b463093d36c1c741d3f0e046c1a6c30d6fefb3e1","source":{"kind":"arxiv","id":"2501.17054","version":1},"attestation_state":"computed","paper":{"title":"Generative diffusion models from a PDE perspective","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"math.PR","authors_text":"(2) Arizona State University, (3) Loyola Marymount University), Fei Cao (1), Justin Le (2), Kimball Johnston (2), S\\'ebastien Motsch (2) ((1) University of Massachusetts Amherst, Thomas Laurent (3)","submitted_at":"2025-01-28T16:29:26Z","abstract_excerpt":"Diffusion models have become the de facto framework for generating new datasets. The core of these models lies in the ability to reverse a diffusion process in time. The goal of this manuscript is to explain, from a PDE perspective, how this method works and how to derive the PDE governing the reverse dynamics as well as to study its solution analytically. By linking forward and reverse dynamics, we show that the reverse process's distribution has its support contained within the original distribution. Consequently, diffusion methods, in their analytical formulation, do not inherently regulari"},"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":"2501.17054","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"math.PR","submitted_at":"2025-01-28T16:29:26Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"aa059440617d606a9420f32cc62f0e913fb65efe9ce13a62410437f74ccdd83f","abstract_canon_sha256":"bb3f9bcd02a861a0d1f0396b62ae1555e958aff6d692c4509325e8f6947ef2bf"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:06:33.557439Z","signature_b64":"+cV6y6xvV9besF1lEQCMU7fMzpH97vdjRfX42eL5zoB90ba3sobaOdJy8iICAxYmqLayKkSQUYvNAEdBAO34CA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6d44425c2aa96d3b4983f1c6b463093d36c1c741d3f0e046c1a6c30d6fefb3e1","last_reissued_at":"2026-07-05T10:06:33.556892Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:06:33.556892Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Generative diffusion models from a PDE perspective","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"math.PR","authors_text":"(2) Arizona State University, (3) Loyola Marymount University), Fei Cao (1), Justin Le (2), Kimball Johnston (2), S\\'ebastien Motsch (2) ((1) University of Massachusetts Amherst, Thomas Laurent (3)","submitted_at":"2025-01-28T16:29:26Z","abstract_excerpt":"Diffusion models have become the de facto framework for generating new datasets. The core of these models lies in the ability to reverse a diffusion process in time. The goal of this manuscript is to explain, from a PDE perspective, how this method works and how to derive the PDE governing the reverse dynamics as well as to study its solution analytically. By linking forward and reverse dynamics, we show that the reverse process's distribution has its support contained within the original distribution. Consequently, diffusion methods, in their analytical formulation, do not inherently regulari"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2501.17054","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/2501.17054/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":"2501.17054","created_at":"2026-07-05T10:06:33.556952+00:00"},{"alias_kind":"arxiv_version","alias_value":"2501.17054v1","created_at":"2026-07-05T10:06:33.556952+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2501.17054","created_at":"2026-07-05T10:06:33.556952+00:00"},{"alias_kind":"pith_short_12","alias_value":"NVCEEXBKVFWT","created_at":"2026-07-05T10:06:33.556952+00:00"},{"alias_kind":"pith_short_16","alias_value":"NVCEEXBKVFWTWSMD","created_at":"2026-07-05T10:06:33.556952+00:00"},{"alias_kind":"pith_short_8","alias_value":"NVCEEXBK","created_at":"2026-07-05T10:06:33.556952+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"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":10,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/NVCEEXBKVFWTWSMD6HDLIYYJHU","json":"https://pith.science/pith/NVCEEXBKVFWTWSMD6HDLIYYJHU.json","graph_json":"https://pith.science/api/pith-number/NVCEEXBKVFWTWSMD6HDLIYYJHU/graph.json","events_json":"https://pith.science/api/pith-number/NVCEEXBKVFWTWSMD6HDLIYYJHU/events.json","paper":"https://pith.science/paper/NVCEEXBK"},"agent_actions":{"view_html":"https://pith.science/pith/NVCEEXBKVFWTWSMD6HDLIYYJHU","download_json":"https://pith.science/pith/NVCEEXBKVFWTWSMD6HDLIYYJHU.json","view_paper":"https://pith.science/paper/NVCEEXBK","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2501.17054&json=true","fetch_graph":"https://pith.science/api/pith-number/NVCEEXBKVFWTWSMD6HDLIYYJHU/graph.json","fetch_events":"https://pith.science/api/pith-number/NVCEEXBKVFWTWSMD6HDLIYYJHU/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/NVCEEXBKVFWTWSMD6HDLIYYJHU/action/timestamp_anchor","attest_storage":"https://pith.science/pith/NVCEEXBKVFWTWSMD6HDLIYYJHU/action/storage_attestation","attest_author":"https://pith.science/pith/NVCEEXBKVFWTWSMD6HDLIYYJHU/action/author_attestation","sign_citation":"https://pith.science/pith/NVCEEXBKVFWTWSMD6HDLIYYJHU/action/citation_signature","submit_replication":"https://pith.science/pith/NVCEEXBKVFWTWSMD6HDLIYYJHU/action/replication_record"}},"created_at":"2026-07-05T10:06:33.556952+00:00","updated_at":"2026-07-05T10:06:33.556952+00:00"}