{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:5VWN7Q5GFTSLKFKB7VPAAWTA4G","short_pith_number":"pith:5VWN7Q5G","schema_version":"1.0","canonical_sha256":"ed6cdfc3a62ce4b51541fd5e005a60e1b6ca975d13603e5fc6ff7070d2cd1429","source":{"kind":"arxiv","id":"2408.12270","version":1},"attestation_state":"computed","paper":{"title":"Variance reduction of diffusion model's gradients with Taylor approximation-based control variate","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Carl Henrik Ek, Jes Frellsen, Michael Riis Andersen, Paul Jeha, Will Grathwohl","submitted_at":"2024-08-22T10:08:34Z","abstract_excerpt":"Score-based models, trained with denoising score matching, are remarkably effective in generating high dimensional data. However, the high variance of their training objective hinders optimisation. We attempt to reduce it with a control variate, derived via a $k$-th order Taylor expansion on the training objective and its gradient. We prove an equivalence between the two and demonstrate empirically the effectiveness of our approach on a low dimensional problem setting; and study its effect on larger problems."},"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":"2408.12270","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2024-08-22T10:08:34Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"f5ebdac786f71d6b6f6a1ae338477e54a3e65f5b8e0f2379cd030204b9ba9eb1","abstract_canon_sha256":"fb18ee0bb1cddb6d379840c8fc7159091b6d75b2f075841fd3598d7dd0389d89"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:58:06.710225Z","signature_b64":"0WJ/HpsynWPo6lEQo0zgIP6xQLKRjZoTz+dFiOerO2QOaX6oT6bxBvAENXpYTG0mdCYvtHaaS6XFYEeeaHakCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ed6cdfc3a62ce4b51541fd5e005a60e1b6ca975d13603e5fc6ff7070d2cd1429","last_reissued_at":"2026-07-05T08:58:06.709819Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:58:06.709819Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Variance reduction of diffusion model's gradients with Taylor approximation-based control variate","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Carl Henrik Ek, Jes Frellsen, Michael Riis Andersen, Paul Jeha, Will Grathwohl","submitted_at":"2024-08-22T10:08:34Z","abstract_excerpt":"Score-based models, trained with denoising score matching, are remarkably effective in generating high dimensional data. However, the high variance of their training objective hinders optimisation. We attempt to reduce it with a control variate, derived via a $k$-th order Taylor expansion on the training objective and its gradient. We prove an equivalence between the two and demonstrate empirically the effectiveness of our approach on a low dimensional problem setting; and study its effect on larger problems."},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2408.12270","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/2408.12270/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":"2408.12270","created_at":"2026-07-05T08:58:06.709879+00:00"},{"alias_kind":"arxiv_version","alias_value":"2408.12270v1","created_at":"2026-07-05T08:58:06.709879+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2408.12270","created_at":"2026-07-05T08:58:06.709879+00:00"},{"alias_kind":"pith_short_12","alias_value":"5VWN7Q5GFTSL","created_at":"2026-07-05T08:58:06.709879+00:00"},{"alias_kind":"pith_short_16","alias_value":"5VWN7Q5GFTSLKFKB","created_at":"2026-07-05T08:58:06.709879+00:00"},{"alias_kind":"pith_short_8","alias_value":"5VWN7Q5G","created_at":"2026-07-05T08:58:06.709879+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2510.03508","citing_title":"D2 Actor Critic: Diffusion Actor Meets Distributional Critic","ref_index":19,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/5VWN7Q5GFTSLKFKB7VPAAWTA4G","json":"https://pith.science/pith/5VWN7Q5GFTSLKFKB7VPAAWTA4G.json","graph_json":"https://pith.science/api/pith-number/5VWN7Q5GFTSLKFKB7VPAAWTA4G/graph.json","events_json":"https://pith.science/api/pith-number/5VWN7Q5GFTSLKFKB7VPAAWTA4G/events.json","paper":"https://pith.science/paper/5VWN7Q5G"},"agent_actions":{"view_html":"https://pith.science/pith/5VWN7Q5GFTSLKFKB7VPAAWTA4G","download_json":"https://pith.science/pith/5VWN7Q5GFTSLKFKB7VPAAWTA4G.json","view_paper":"https://pith.science/paper/5VWN7Q5G","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2408.12270&json=true","fetch_graph":"https://pith.science/api/pith-number/5VWN7Q5GFTSLKFKB7VPAAWTA4G/graph.json","fetch_events":"https://pith.science/api/pith-number/5VWN7Q5GFTSLKFKB7VPAAWTA4G/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/5VWN7Q5GFTSLKFKB7VPAAWTA4G/action/timestamp_anchor","attest_storage":"https://pith.science/pith/5VWN7Q5GFTSLKFKB7VPAAWTA4G/action/storage_attestation","attest_author":"https://pith.science/pith/5VWN7Q5GFTSLKFKB7VPAAWTA4G/action/author_attestation","sign_citation":"https://pith.science/pith/5VWN7Q5GFTSLKFKB7VPAAWTA4G/action/citation_signature","submit_replication":"https://pith.science/pith/5VWN7Q5GFTSLKFKB7VPAAWTA4G/action/replication_record"}},"created_at":"2026-07-05T08:58:06.709879+00:00","updated_at":"2026-07-05T08:58:06.709879+00:00"}