{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:MC732QXBWHFSVBNDKLKZXX53OF","short_pith_number":"pith:MC732QXB","schema_version":"1.0","canonical_sha256":"60bfbd42e1b1cb2a85a352d59bdfbb716c7798ee80afd6217dd6d69ecf294afe","source":{"kind":"arxiv","id":"2104.02600","version":2},"attestation_state":"computed","paper":{"title":"Noise Estimation for Generative Diffusion Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CV"],"primary_cat":"cs.LG","authors_text":"Eliya Nachmani, Lior Wolf, Robin San-Roman","submitted_at":"2021-04-06T15:46:16Z","abstract_excerpt":"Generative diffusion models have emerged as leading models in speech and image generation. However, in order to perform well with a small number of denoising steps, a costly tuning of the set of noise parameters is needed. In this work, we present a simple and versatile learning scheme that can step-by-step adjust those noise parameters, for any given number of steps, while the previous work needs to retune for each number separately. Furthermore, without modifying the weights of the diffusion model, we are able to significantly improve the synthesis results, for a small number of steps. Our a"},"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":"2104.02600","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2021-04-06T15:46:16Z","cross_cats_sorted":["cs.CV"],"title_canon_sha256":"fa046cc48c51912616109ed11f7361266d45deac3c1e6850548e3dd93ba45b52","abstract_canon_sha256":"b122785ea5db31fb8128adeb49809fcf93de4d1321d92960648bb00e69b05f2d"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T03:13:26.668659Z","signature_b64":"gt8IOl3LPVJgTsMTl46yFslwR6V29a8E37uHUbwrfO82U5I3M3ZedPmbhQDTYfLCKD/7W9zDzBKGfuCmzpoPDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"60bfbd42e1b1cb2a85a352d59bdfbb716c7798ee80afd6217dd6d69ecf294afe","last_reissued_at":"2026-07-05T03:13:26.668238Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T03:13:26.668238Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Noise Estimation for Generative Diffusion Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CV"],"primary_cat":"cs.LG","authors_text":"Eliya Nachmani, Lior Wolf, Robin San-Roman","submitted_at":"2021-04-06T15:46:16Z","abstract_excerpt":"Generative diffusion models have emerged as leading models in speech and image generation. However, in order to perform well with a small number of denoising steps, a costly tuning of the set of noise parameters is needed. In this work, we present a simple and versatile learning scheme that can step-by-step adjust those noise parameters, for any given number of steps, while the previous work needs to retune for each number separately. Furthermore, without modifying the weights of the diffusion model, we are able to significantly improve the synthesis results, for a small number of steps. Our a"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2104.02600","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/2104.02600/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":"2104.02600","created_at":"2026-07-05T03:13:26.668298+00:00"},{"alias_kind":"arxiv_version","alias_value":"2104.02600v2","created_at":"2026-07-05T03:13:26.668298+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2104.02600","created_at":"2026-07-05T03:13:26.668298+00:00"},{"alias_kind":"pith_short_12","alias_value":"MC732QXBWHFS","created_at":"2026-07-05T03:13:26.668298+00:00"},{"alias_kind":"pith_short_16","alias_value":"MC732QXBWHFSVBND","created_at":"2026-07-05T03:13:26.668298+00:00"},{"alias_kind":"pith_short_8","alias_value":"MC732QXB","created_at":"2026-07-05T03:13:26.668298+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":6,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2311.04938","citing_title":"Improved DDIM Sampling with Moment Matching Gaussian Mixtures","ref_index":20,"is_internal_anchor":false},{"citing_arxiv_id":"2501.02576","citing_title":"DepthMaster: Taming Diffusion Models for Monocular Depth Estimation","ref_index":33,"is_internal_anchor":false},{"citing_arxiv_id":"2509.01629","citing_title":"Lipschitz-Guided Design of Interpolation Schedules in Generative Models","ref_index":35,"is_internal_anchor":false},{"citing_arxiv_id":"2211.01095","citing_title":"DPM-Solver++: Fast Solver for Guided Sampling of Diffusion Probabilistic Models","ref_index":11,"is_internal_anchor":false},{"citing_arxiv_id":"2112.10752","citing_title":"High-Resolution Image Synthesis with Latent Diffusion Models","ref_index":75,"is_internal_anchor":false},{"citing_arxiv_id":"2202.00512","citing_title":"Progressive Distillation for Fast Sampling of Diffusion Models","ref_index":17,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/MC732QXBWHFSVBNDKLKZXX53OF","json":"https://pith.science/pith/MC732QXBWHFSVBNDKLKZXX53OF.json","graph_json":"https://pith.science/api/pith-number/MC732QXBWHFSVBNDKLKZXX53OF/graph.json","events_json":"https://pith.science/api/pith-number/MC732QXBWHFSVBNDKLKZXX53OF/events.json","paper":"https://pith.science/paper/MC732QXB"},"agent_actions":{"view_html":"https://pith.science/pith/MC732QXBWHFSVBNDKLKZXX53OF","download_json":"https://pith.science/pith/MC732QXBWHFSVBNDKLKZXX53OF.json","view_paper":"https://pith.science/paper/MC732QXB","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2104.02600&json=true","fetch_graph":"https://pith.science/api/pith-number/MC732QXBWHFSVBNDKLKZXX53OF/graph.json","fetch_events":"https://pith.science/api/pith-number/MC732QXBWHFSVBNDKLKZXX53OF/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/MC732QXBWHFSVBNDKLKZXX53OF/action/timestamp_anchor","attest_storage":"https://pith.science/pith/MC732QXBWHFSVBNDKLKZXX53OF/action/storage_attestation","attest_author":"https://pith.science/pith/MC732QXBWHFSVBNDKLKZXX53OF/action/author_attestation","sign_citation":"https://pith.science/pith/MC732QXBWHFSVBNDKLKZXX53OF/action/citation_signature","submit_replication":"https://pith.science/pith/MC732QXBWHFSVBNDKLKZXX53OF/action/replication_record"}},"created_at":"2026-07-05T03:13:26.668298+00:00","updated_at":"2026-07-05T03:13:26.668298+00:00"}