{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:SRTJYK5JXJNKOVRHGDH24NHBNP","short_pith_number":"pith:SRTJYK5J","schema_version":"1.0","canonical_sha256":"94669c2ba9ba5aa7562730cfae34e16bff7062c66c46d733c82fd9e44c8323b9","source":{"kind":"arxiv","id":"2303.03384","version":1},"attestation_state":"computed","paper":{"title":"Restoration-Degradation Beyond Linear Diffusions: A Non-Asymptotic Analysis For DDIM-Type Samplers","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["math.ST","stat.ML","stat.TH"],"primary_cat":"cs.LG","authors_text":"Alexandros G. Dimakis, Giannis Daras, Sitan Chen","submitted_at":"2023-03-06T18:59:19Z","abstract_excerpt":"We develop a framework for non-asymptotic analysis of deterministic samplers used for diffusion generative modeling. Several recent works have analyzed stochastic samplers using tools like Girsanov's theorem and a chain rule variant of the interpolation argument. Unfortunately, these techniques give vacuous bounds when applied to deterministic samplers. We give a new operational interpretation for deterministic sampling by showing that one step along the probability flow ODE can be expressed as two steps: 1) a restoration step that runs gradient ascent on the conditional log-likelihood at some"},"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":"2303.03384","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2023-03-06T18:59:19Z","cross_cats_sorted":["math.ST","stat.ML","stat.TH"],"title_canon_sha256":"ab93ccb51d8f7d24bb38d2880fc135cc2460950a6ad550db6d4d375aedddb11b","abstract_canon_sha256":"cb111b658700b041590dd2057a3f177eed151caeabb28c47c90772bf450f96ee"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:48:29.067788Z","signature_b64":"duvDggLFiX/H968Tom4ODCxnkcsoPNSgjruOcpmlIAs1DillnXdacu78YcVdoWagKij2geci6teTi0G51TqCAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"94669c2ba9ba5aa7562730cfae34e16bff7062c66c46d733c82fd9e44c8323b9","last_reissued_at":"2026-07-05T05:48:29.067288Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:48:29.067288Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Restoration-Degradation Beyond Linear Diffusions: A Non-Asymptotic Analysis For DDIM-Type Samplers","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["math.ST","stat.ML","stat.TH"],"primary_cat":"cs.LG","authors_text":"Alexandros G. Dimakis, Giannis Daras, Sitan Chen","submitted_at":"2023-03-06T18:59:19Z","abstract_excerpt":"We develop a framework for non-asymptotic analysis of deterministic samplers used for diffusion generative modeling. Several recent works have analyzed stochastic samplers using tools like Girsanov's theorem and a chain rule variant of the interpolation argument. Unfortunately, these techniques give vacuous bounds when applied to deterministic samplers. We give a new operational interpretation for deterministic sampling by showing that one step along the probability flow ODE can be expressed as two steps: 1) a restoration step that runs gradient ascent on the conditional log-likelihood at some"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2303.03384","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/2303.03384/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":"2303.03384","created_at":"2026-07-05T05:48:29.067362+00:00"},{"alias_kind":"arxiv_version","alias_value":"2303.03384v1","created_at":"2026-07-05T05:48:29.067362+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2303.03384","created_at":"2026-07-05T05:48:29.067362+00:00"},{"alias_kind":"pith_short_12","alias_value":"SRTJYK5JXJNK","created_at":"2026-07-05T05:48:29.067362+00:00"},{"alias_kind":"pith_short_16","alias_value":"SRTJYK5JXJNKOVRH","created_at":"2026-07-05T05:48:29.067362+00:00"},{"alias_kind":"pith_short_8","alias_value":"SRTJYK5J","created_at":"2026-07-05T05:48:29.067362+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2506.24042","citing_title":"Faster Diffusion Models via Higher-Order Approximation","ref_index":8,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/SRTJYK5JXJNKOVRHGDH24NHBNP","json":"https://pith.science/pith/SRTJYK5JXJNKOVRHGDH24NHBNP.json","graph_json":"https://pith.science/api/pith-number/SRTJYK5JXJNKOVRHGDH24NHBNP/graph.json","events_json":"https://pith.science/api/pith-number/SRTJYK5JXJNKOVRHGDH24NHBNP/events.json","paper":"https://pith.science/paper/SRTJYK5J"},"agent_actions":{"view_html":"https://pith.science/pith/SRTJYK5JXJNKOVRHGDH24NHBNP","download_json":"https://pith.science/pith/SRTJYK5JXJNKOVRHGDH24NHBNP.json","view_paper":"https://pith.science/paper/SRTJYK5J","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2303.03384&json=true","fetch_graph":"https://pith.science/api/pith-number/SRTJYK5JXJNKOVRHGDH24NHBNP/graph.json","fetch_events":"https://pith.science/api/pith-number/SRTJYK5JXJNKOVRHGDH24NHBNP/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/SRTJYK5JXJNKOVRHGDH24NHBNP/action/timestamp_anchor","attest_storage":"https://pith.science/pith/SRTJYK5JXJNKOVRHGDH24NHBNP/action/storage_attestation","attest_author":"https://pith.science/pith/SRTJYK5JXJNKOVRHGDH24NHBNP/action/author_attestation","sign_citation":"https://pith.science/pith/SRTJYK5JXJNKOVRHGDH24NHBNP/action/citation_signature","submit_replication":"https://pith.science/pith/SRTJYK5JXJNKOVRHGDH24NHBNP/action/replication_record"}},"created_at":"2026-07-05T05:48:29.067362+00:00","updated_at":"2026-07-05T05:48:29.067362+00:00"}