{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:W4T6BLLZGSA75R2JLUPQSJXIAY","short_pith_number":"pith:W4T6BLLZ","schema_version":"1.0","canonical_sha256":"b727e0ad793481fec7495d1f0926e8061a0ee7cb40aab2f9e6e5bdb1763a445a","source":{"kind":"arxiv","id":"2402.02149","version":2},"attestation_state":"computed","paper":{"title":"Improving Diffusion Models for Inverse Problems Using Optimal Posterior Covariance","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CV","authors_text":"Chenglin Li, Hongkai Xiong, Junni Zou, Nuoqian Xiao, Wenrui Dai, Xinyu Peng, Ziyang Zheng","submitted_at":"2024-02-03T13:35:39Z","abstract_excerpt":"Recent diffusion models provide a promising zero-shot solution to noisy linear inverse problems without retraining for specific inverse problems. In this paper, we reveal that recent methods can be uniformly interpreted as employing a Gaussian approximation with hand-crafted isotropic covariance for the intractable denoising posterior to approximate the conditional posterior mean. Inspired by this finding, we propose to improve recent methods by using more principled covariance determined by maximum likelihood estimation. To achieve posterior covariance optimization without retraining, we prov"},"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":"2402.02149","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2024-02-03T13:35:39Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"940307ffb8ed180d6c2a1605db07c229409fae261b9f57c9914a222f61dc5d73","abstract_canon_sha256":"89acd01740c20537b1a9e4d0f9469c37e01502aec447a171c30b8943839577a1"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:26:18.643953Z","signature_b64":"wgdi/4o7mEY1jYKD4G7uFS8Wkh7eXxVO+AH1zL3KqeslPTmrFBb9F00SFLypVlCDxvagIRkO0ugbLxdaOkSFAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b727e0ad793481fec7495d1f0926e8061a0ee7cb40aab2f9e6e5bdb1763a445a","last_reissued_at":"2026-07-05T08:26:18.643400Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:26:18.643400Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Improving Diffusion Models for Inverse Problems Using Optimal Posterior Covariance","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CV","authors_text":"Chenglin Li, Hongkai Xiong, Junni Zou, Nuoqian Xiao, Wenrui Dai, Xinyu Peng, Ziyang Zheng","submitted_at":"2024-02-03T13:35:39Z","abstract_excerpt":"Recent diffusion models provide a promising zero-shot solution to noisy linear inverse problems without retraining for specific inverse problems. In this paper, we reveal that recent methods can be uniformly interpreted as employing a Gaussian approximation with hand-crafted isotropic covariance for the intractable denoising posterior to approximate the conditional posterior mean. Inspired by this finding, we propose to improve recent methods by using more principled covariance determined by maximum likelihood estimation. To achieve posterior covariance optimization without retraining, we prov"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2402.02149","kind":"arxiv","version":2},"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/2402.02149/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":"2402.02149","created_at":"2026-07-05T08:26:18.643466+00:00"},{"alias_kind":"arxiv_version","alias_value":"2402.02149v2","created_at":"2026-07-05T08:26:18.643466+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2402.02149","created_at":"2026-07-05T08:26:18.643466+00:00"},{"alias_kind":"pith_short_12","alias_value":"W4T6BLLZGSA7","created_at":"2026-07-05T08:26:18.643466+00:00"},{"alias_kind":"pith_short_16","alias_value":"W4T6BLLZGSA75R2J","created_at":"2026-07-05T08:26:18.643466+00:00"},{"alias_kind":"pith_short_8","alias_value":"W4T6BLLZ","created_at":"2026-07-05T08:26:18.643466+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2602.20549","citing_title":"Sample-efficient evidence estimation of score based priors for model selection","ref_index":13,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/W4T6BLLZGSA75R2JLUPQSJXIAY","json":"https://pith.science/pith/W4T6BLLZGSA75R2JLUPQSJXIAY.json","graph_json":"https://pith.science/api/pith-number/W4T6BLLZGSA75R2JLUPQSJXIAY/graph.json","events_json":"https://pith.science/api/pith-number/W4T6BLLZGSA75R2JLUPQSJXIAY/events.json","paper":"https://pith.science/paper/W4T6BLLZ"},"agent_actions":{"view_html":"https://pith.science/pith/W4T6BLLZGSA75R2JLUPQSJXIAY","download_json":"https://pith.science/pith/W4T6BLLZGSA75R2JLUPQSJXIAY.json","view_paper":"https://pith.science/paper/W4T6BLLZ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2402.02149&json=true","fetch_graph":"https://pith.science/api/pith-number/W4T6BLLZGSA75R2JLUPQSJXIAY/graph.json","fetch_events":"https://pith.science/api/pith-number/W4T6BLLZGSA75R2JLUPQSJXIAY/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/W4T6BLLZGSA75R2JLUPQSJXIAY/action/timestamp_anchor","attest_storage":"https://pith.science/pith/W4T6BLLZGSA75R2JLUPQSJXIAY/action/storage_attestation","attest_author":"https://pith.science/pith/W4T6BLLZGSA75R2JLUPQSJXIAY/action/author_attestation","sign_citation":"https://pith.science/pith/W4T6BLLZGSA75R2JLUPQSJXIAY/action/citation_signature","submit_replication":"https://pith.science/pith/W4T6BLLZGSA75R2JLUPQSJXIAY/action/replication_record"}},"created_at":"2026-07-05T08:26:18.643466+00:00","updated_at":"2026-07-05T08:26:18.643466+00:00"}