{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:LIWMNAPUKD3J2ZQXZGAVSYLRA5","short_pith_number":"pith:LIWMNAPU","schema_version":"1.0","canonical_sha256":"5a2cc681f450f69d6617c981596171077c59f1382e3775745702d884c798d37b","source":{"kind":"arxiv","id":"2307.10711","version":3},"attestation_state":"computed","paper":{"title":"AdjointDPM: Adjoint Sensitivity Method for Gradient Backpropagation of Diffusion Probabilistic Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CV","authors_text":"Hanshu Yan, Jiachun Pan, Jiashi Feng, Jun Hao Liew, Vincent Y. F. Tan","submitted_at":"2023-07-20T09:06:21Z","abstract_excerpt":"Existing customization methods require access to multiple reference examples to align pre-trained diffusion probabilistic models (DPMs) with user-provided concepts. This paper aims to address the challenge of DPM customization when the only available supervision is a differentiable metric defined on the generated contents. Since the sampling procedure of DPMs involves recursive calls to the denoising UNet, na\\\"ive gradient backpropagation requires storing the intermediate states of all iterations, resulting in extremely high memory consumption. To overcome this issue, we propose a novel method"},"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":"2307.10711","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2023-07-20T09:06:21Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"28633561f1f8243dddaaba6aa66a6b9ca75b366df8a758e446c978a5c8be7d9f","abstract_canon_sha256":"87c4cc6fd14c84fd13da878bc87b3569c5f247fe47469e42027b53fc64da9a9d"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:58:22.153852Z","signature_b64":"Da7XOOrqPQgNChXAgNf7hXLdNR36llmYbnDASaNS0SPx1LR7sP+gjYQf0DCIUBxNbtFWcQ9dDeTp+TGgfiltAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5a2cc681f450f69d6617c981596171077c59f1382e3775745702d884c798d37b","last_reissued_at":"2026-07-05T07:58:22.153311Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:58:22.153311Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"AdjointDPM: Adjoint Sensitivity Method for Gradient Backpropagation of Diffusion Probabilistic Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CV","authors_text":"Hanshu Yan, Jiachun Pan, Jiashi Feng, Jun Hao Liew, Vincent Y. F. Tan","submitted_at":"2023-07-20T09:06:21Z","abstract_excerpt":"Existing customization methods require access to multiple reference examples to align pre-trained diffusion probabilistic models (DPMs) with user-provided concepts. This paper aims to address the challenge of DPM customization when the only available supervision is a differentiable metric defined on the generated contents. Since the sampling procedure of DPMs involves recursive calls to the denoising UNet, na\\\"ive gradient backpropagation requires storing the intermediate states of all iterations, resulting in extremely high memory consumption. To overcome this issue, we propose a novel method"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2307.10711","kind":"arxiv","version":3},"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/2307.10711/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":"2307.10711","created_at":"2026-07-05T07:58:22.153392+00:00"},{"alias_kind":"arxiv_version","alias_value":"2307.10711v3","created_at":"2026-07-05T07:58:22.153392+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2307.10711","created_at":"2026-07-05T07:58:22.153392+00:00"},{"alias_kind":"pith_short_12","alias_value":"LIWMNAPUKD3J","created_at":"2026-07-05T07:58:22.153392+00:00"},{"alias_kind":"pith_short_16","alias_value":"LIWMNAPUKD3J2ZQX","created_at":"2026-07-05T07:58:22.153392+00:00"},{"alias_kind":"pith_short_8","alias_value":"LIWMNAPU","created_at":"2026-07-05T07:58:22.153392+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2505.23264","citing_title":"Efficiently Access Diffusion Fisher: Within the Outer Product Span Space","ref_index":39,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/LIWMNAPUKD3J2ZQXZGAVSYLRA5","json":"https://pith.science/pith/LIWMNAPUKD3J2ZQXZGAVSYLRA5.json","graph_json":"https://pith.science/api/pith-number/LIWMNAPUKD3J2ZQXZGAVSYLRA5/graph.json","events_json":"https://pith.science/api/pith-number/LIWMNAPUKD3J2ZQXZGAVSYLRA5/events.json","paper":"https://pith.science/paper/LIWMNAPU"},"agent_actions":{"view_html":"https://pith.science/pith/LIWMNAPUKD3J2ZQXZGAVSYLRA5","download_json":"https://pith.science/pith/LIWMNAPUKD3J2ZQXZGAVSYLRA5.json","view_paper":"https://pith.science/paper/LIWMNAPU","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2307.10711&json=true","fetch_graph":"https://pith.science/api/pith-number/LIWMNAPUKD3J2ZQXZGAVSYLRA5/graph.json","fetch_events":"https://pith.science/api/pith-number/LIWMNAPUKD3J2ZQXZGAVSYLRA5/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/LIWMNAPUKD3J2ZQXZGAVSYLRA5/action/timestamp_anchor","attest_storage":"https://pith.science/pith/LIWMNAPUKD3J2ZQXZGAVSYLRA5/action/storage_attestation","attest_author":"https://pith.science/pith/LIWMNAPUKD3J2ZQXZGAVSYLRA5/action/author_attestation","sign_citation":"https://pith.science/pith/LIWMNAPUKD3J2ZQXZGAVSYLRA5/action/citation_signature","submit_replication":"https://pith.science/pith/LIWMNAPUKD3J2ZQXZGAVSYLRA5/action/replication_record"}},"created_at":"2026-07-05T07:58:22.153392+00:00","updated_at":"2026-07-05T07:58:22.153392+00:00"}