{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:KN4DGT4VC6QQQFOLKZA3LVKW47","short_pith_number":"pith:KN4DGT4V","schema_version":"1.0","canonical_sha256":"5378334f9517a10815cb5641b5d556e7f1534f24a08534ae16fbd196398f25c4","source":{"kind":"arxiv","id":"2307.01178","version":1},"attestation_state":"computed","paper":{"title":"Learning Mixtures of Gaussians Using the DDPM Objective","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG","stat.ML"],"primary_cat":"cs.DS","authors_text":"Adam Klivans, Kulin Shah, Sitan Chen","submitted_at":"2023-07-03T17:44:22Z","abstract_excerpt":"Recent works have shown that diffusion models can learn essentially any distribution provided one can perform score estimation. Yet it remains poorly understood under what settings score estimation is possible, let alone when practical gradient-based algorithms for this task can provably succeed.\n  In this work, we give the first provably efficient results along these lines for one of the most fundamental distribution families, Gaussian mixture models. We prove that gradient descent on the denoising diffusion probabilistic model (DDPM) objective can efficiently recover the ground truth paramet"},"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":"2307.01178","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.DS","submitted_at":"2023-07-03T17:44:22Z","cross_cats_sorted":["cs.LG","stat.ML"],"title_canon_sha256":"c8d0890b88d116d9a583a1f36a5567c70451eafaefab403c1633eaec81a2da8b","abstract_canon_sha256":"a74912a7ca8be92c74f6c494787936216cc5a4a32eababe1485436d2461cf347"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:27:04.783020Z","signature_b64":"XiInSJb0XsVKPbIYFPTlQ9cdG336XjJWbKzacdUZCVlVNyGHqR15B5x6QVITnAIndk0cfBAWcaITdDn3QSa7AQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5378334f9517a10815cb5641b5d556e7f1534f24a08534ae16fbd196398f25c4","last_reissued_at":"2026-07-05T06:27:04.782493Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:27:04.782493Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Learning Mixtures of Gaussians Using the DDPM Objective","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG","stat.ML"],"primary_cat":"cs.DS","authors_text":"Adam Klivans, Kulin Shah, Sitan Chen","submitted_at":"2023-07-03T17:44:22Z","abstract_excerpt":"Recent works have shown that diffusion models can learn essentially any distribution provided one can perform score estimation. Yet it remains poorly understood under what settings score estimation is possible, let alone when practical gradient-based algorithms for this task can provably succeed.\n  In this work, we give the first provably efficient results along these lines for one of the most fundamental distribution families, Gaussian mixture models. We prove that gradient descent on the denoising diffusion probabilistic model (DDPM) objective can efficiently recover the ground truth paramet"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2307.01178","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/2307.01178/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":"2307.01178","created_at":"2026-07-05T06:27:04.782550+00:00"},{"alias_kind":"arxiv_version","alias_value":"2307.01178v1","created_at":"2026-07-05T06:27:04.782550+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2307.01178","created_at":"2026-07-05T06:27:04.782550+00:00"},{"alias_kind":"pith_short_12","alias_value":"KN4DGT4VC6QQ","created_at":"2026-07-05T06:27:04.782550+00:00"},{"alias_kind":"pith_short_16","alias_value":"KN4DGT4VC6QQQFOL","created_at":"2026-07-05T06:27:04.782550+00:00"},{"alias_kind":"pith_short_8","alias_value":"KN4DGT4V","created_at":"2026-07-05T06:27:04.782550+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.20299","citing_title":"Statistical Properties of Training & Generalization","ref_index":265,"is_internal_anchor":false},{"citing_arxiv_id":"2606.20299","citing_title":"Statistical Properties of Training & Generalization","ref_index":265,"is_internal_anchor":false},{"citing_arxiv_id":"2603.12901","citing_title":"A theory of learning data statistics in diffusion models, from easy to hard","ref_index":36,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/KN4DGT4VC6QQQFOLKZA3LVKW47","json":"https://pith.science/pith/KN4DGT4VC6QQQFOLKZA3LVKW47.json","graph_json":"https://pith.science/api/pith-number/KN4DGT4VC6QQQFOLKZA3LVKW47/graph.json","events_json":"https://pith.science/api/pith-number/KN4DGT4VC6QQQFOLKZA3LVKW47/events.json","paper":"https://pith.science/paper/KN4DGT4V"},"agent_actions":{"view_html":"https://pith.science/pith/KN4DGT4VC6QQQFOLKZA3LVKW47","download_json":"https://pith.science/pith/KN4DGT4VC6QQQFOLKZA3LVKW47.json","view_paper":"https://pith.science/paper/KN4DGT4V","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2307.01178&json=true","fetch_graph":"https://pith.science/api/pith-number/KN4DGT4VC6QQQFOLKZA3LVKW47/graph.json","fetch_events":"https://pith.science/api/pith-number/KN4DGT4VC6QQQFOLKZA3LVKW47/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/KN4DGT4VC6QQQFOLKZA3LVKW47/action/timestamp_anchor","attest_storage":"https://pith.science/pith/KN4DGT4VC6QQQFOLKZA3LVKW47/action/storage_attestation","attest_author":"https://pith.science/pith/KN4DGT4VC6QQQFOLKZA3LVKW47/action/author_attestation","sign_citation":"https://pith.science/pith/KN4DGT4VC6QQQFOLKZA3LVKW47/action/citation_signature","submit_replication":"https://pith.science/pith/KN4DGT4VC6QQQFOLKZA3LVKW47/action/replication_record"}},"created_at":"2026-07-05T06:27:04.782550+00:00","updated_at":"2026-07-05T06:27:04.782550+00:00"}