{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:NQD3FY53FHBHXO2UQXY6HGK4CO","short_pith_number":"pith:NQD3FY53","schema_version":"1.0","canonical_sha256":"6c07b2e3bb29c27bbb5485f1e3995c138fbf4d0cb6f728d768e53628c9b79ac8","source":{"kind":"arxiv","id":"2206.07309","version":1},"attestation_state":"computed","paper":{"title":"Estimating the Optimal Covariance with Imperfect Mean in Diffusion Probabilistic Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Bo Zhang, Chongxuan Li, Fan Bao, Jiacheng Sun, Jun Zhu","submitted_at":"2022-06-15T05:42:48Z","abstract_excerpt":"Diffusion probabilistic models (DPMs) are a class of powerful deep generative models (DGMs). Despite their success, the iterative generation process over the full timesteps is much less efficient than other DGMs such as GANs. Thus, the generation performance on a subset of timesteps is crucial, which is greatly influenced by the covariance design in DPMs. In this work, we consider diagonal and full covariances to improve the expressive power of DPMs. We derive the optimal result for such covariances, and then correct it when the mean of DPMs is imperfect. Both the optimal and the corrected one"},"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":"2206.07309","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2022-06-15T05:42:48Z","cross_cats_sorted":[],"title_canon_sha256":"f82abfde564b23eaed86bd6e30e4168737d7c624557958e46557a3c92eeb4918","abstract_canon_sha256":"930fd10e1daadb37a91bf518bf97e2486cda3eb3b7467ffb4e24609e8ef4d319"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T04:32:03.304039Z","signature_b64":"9+l71wsbJOy3ujTzTbHcqDv24kw/1Lmyu0rzfn/Kum4UITVrImH3q9BVUZS+TammNEbDINMFMXLYn3n5L5O/Dw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6c07b2e3bb29c27bbb5485f1e3995c138fbf4d0cb6f728d768e53628c9b79ac8","last_reissued_at":"2026-07-05T04:32:03.303573Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T04:32:03.303573Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Estimating the Optimal Covariance with Imperfect Mean in Diffusion Probabilistic Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Bo Zhang, Chongxuan Li, Fan Bao, Jiacheng Sun, Jun Zhu","submitted_at":"2022-06-15T05:42:48Z","abstract_excerpt":"Diffusion probabilistic models (DPMs) are a class of powerful deep generative models (DGMs). Despite their success, the iterative generation process over the full timesteps is much less efficient than other DGMs such as GANs. Thus, the generation performance on a subset of timesteps is crucial, which is greatly influenced by the covariance design in DPMs. In this work, we consider diagonal and full covariances to improve the expressive power of DPMs. We derive the optimal result for such covariances, and then correct it when the mean of DPMs is imperfect. Both the optimal and the corrected one"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2206.07309","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/2206.07309/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":"2206.07309","created_at":"2026-07-05T04:32:03.303631+00:00"},{"alias_kind":"arxiv_version","alias_value":"2206.07309v1","created_at":"2026-07-05T04:32:03.303631+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2206.07309","created_at":"2026-07-05T04:32:03.303631+00:00"},{"alias_kind":"pith_short_12","alias_value":"NQD3FY53FHBH","created_at":"2026-07-05T04:32:03.303631+00:00"},{"alias_kind":"pith_short_16","alias_value":"NQD3FY53FHBHXO2U","created_at":"2026-07-05T04:32:03.303631+00:00"},{"alias_kind":"pith_short_8","alias_value":"NQD3FY53","created_at":"2026-07-05T04:32:03.303631+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.22723","citing_title":"The Value of Covariance Matching in Gaussian DDPMs and the Lanczos Sampler","ref_index":1,"is_internal_anchor":false},{"citing_arxiv_id":"2211.01095","citing_title":"DPM-Solver++: Fast Solver for Guided Sampling of Diffusion Probabilistic Models","ref_index":1,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/NQD3FY53FHBHXO2UQXY6HGK4CO","json":"https://pith.science/pith/NQD3FY53FHBHXO2UQXY6HGK4CO.json","graph_json":"https://pith.science/api/pith-number/NQD3FY53FHBHXO2UQXY6HGK4CO/graph.json","events_json":"https://pith.science/api/pith-number/NQD3FY53FHBHXO2UQXY6HGK4CO/events.json","paper":"https://pith.science/paper/NQD3FY53"},"agent_actions":{"view_html":"https://pith.science/pith/NQD3FY53FHBHXO2UQXY6HGK4CO","download_json":"https://pith.science/pith/NQD3FY53FHBHXO2UQXY6HGK4CO.json","view_paper":"https://pith.science/paper/NQD3FY53","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2206.07309&json=true","fetch_graph":"https://pith.science/api/pith-number/NQD3FY53FHBHXO2UQXY6HGK4CO/graph.json","fetch_events":"https://pith.science/api/pith-number/NQD3FY53FHBHXO2UQXY6HGK4CO/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/NQD3FY53FHBHXO2UQXY6HGK4CO/action/timestamp_anchor","attest_storage":"https://pith.science/pith/NQD3FY53FHBHXO2UQXY6HGK4CO/action/storage_attestation","attest_author":"https://pith.science/pith/NQD3FY53FHBHXO2UQXY6HGK4CO/action/author_attestation","sign_citation":"https://pith.science/pith/NQD3FY53FHBHXO2UQXY6HGK4CO/action/citation_signature","submit_replication":"https://pith.science/pith/NQD3FY53FHBHXO2UQXY6HGK4CO/action/replication_record"}},"created_at":"2026-07-05T04:32:03.303631+00:00","updated_at":"2026-07-05T04:32:03.303631+00:00"}