{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:SFCUMWGVZBMPBDTJLC5LLN3QXO","short_pith_number":"pith:SFCUMWGV","schema_version":"1.0","canonical_sha256":"91454658d5c858f08e6958bab5b770bbb1c15a1d66c1091279ea78fbe145a717","source":{"kind":"arxiv","id":"2507.03756","version":1},"attestation_state":"computed","paper":{"title":"Implicit Regularisation in Diffusion Models: An Algorithm-Dependent Generalisation Analysis","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG","math.ST","stat.TH"],"primary_cat":"stat.ML","authors_text":"Arnaud Doucet, George Deligiannidis, Patrick Rebeschini, Tyler Farghly","submitted_at":"2025-07-04T18:07:06Z","abstract_excerpt":"The success of denoising diffusion models raises important questions regarding their generalisation behaviour, particularly in high-dimensional settings. Notably, it has been shown that when training and sampling are performed perfectly, these models memorise training data -- implying that some form of regularisation is essential for generalisation. Existing theoretical analyses primarily rely on algorithm-independent techniques such as uniform convergence, heavily utilising model structure to obtain generalisation bounds. In this work, we instead leverage the algorithmic aspects that promote "},"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":"2507.03756","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"stat.ML","submitted_at":"2025-07-04T18:07:06Z","cross_cats_sorted":["cs.LG","math.ST","stat.TH"],"title_canon_sha256":"8d1bf183f236230d9d8e8cc5ade3e9585de6cd27898a5a29b40221d9dc79c0bf","abstract_canon_sha256":"1aaaa6cc85e3ab93689efa7c551f590814cdce59e391017e6d6ff4e3253b33cd"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:32:04.586701Z","signature_b64":"i96RvoBen44mB99nwBEsXdOahlFJBxykT98HPfsECgPLRNMTg5HBihtjlTBlzhTDlDg1wnTjuUoxYk/GlUBqAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"91454658d5c858f08e6958bab5b770bbb1c15a1d66c1091279ea78fbe145a717","last_reissued_at":"2026-07-05T11:32:04.586216Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:32:04.586216Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Implicit Regularisation in Diffusion Models: An Algorithm-Dependent Generalisation Analysis","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG","math.ST","stat.TH"],"primary_cat":"stat.ML","authors_text":"Arnaud Doucet, George Deligiannidis, Patrick Rebeschini, Tyler Farghly","submitted_at":"2025-07-04T18:07:06Z","abstract_excerpt":"The success of denoising diffusion models raises important questions regarding their generalisation behaviour, particularly in high-dimensional settings. Notably, it has been shown that when training and sampling are performed perfectly, these models memorise training data -- implying that some form of regularisation is essential for generalisation. Existing theoretical analyses primarily rely on algorithm-independent techniques such as uniform convergence, heavily utilising model structure to obtain generalisation bounds. In this work, we instead leverage the algorithmic aspects that promote "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2507.03756","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/2507.03756/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":"2507.03756","created_at":"2026-07-05T11:32:04.586268+00:00"},{"alias_kind":"arxiv_version","alias_value":"2507.03756v1","created_at":"2026-07-05T11:32:04.586268+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2507.03756","created_at":"2026-07-05T11:32:04.586268+00:00"},{"alias_kind":"pith_short_12","alias_value":"SFCUMWGVZBMP","created_at":"2026-07-05T11:32:04.586268+00:00"},{"alias_kind":"pith_short_16","alias_value":"SFCUMWGVZBMPBDTJ","created_at":"2026-07-05T11:32:04.586268+00:00"},{"alias_kind":"pith_short_8","alias_value":"SFCUMWGV","created_at":"2026-07-05T11:32:04.586268+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2603.13419","citing_title":"Diffusion Models Memorize in Training -- and Generalize in Inference","ref_index":19,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/SFCUMWGVZBMPBDTJLC5LLN3QXO","json":"https://pith.science/pith/SFCUMWGVZBMPBDTJLC5LLN3QXO.json","graph_json":"https://pith.science/api/pith-number/SFCUMWGVZBMPBDTJLC5LLN3QXO/graph.json","events_json":"https://pith.science/api/pith-number/SFCUMWGVZBMPBDTJLC5LLN3QXO/events.json","paper":"https://pith.science/paper/SFCUMWGV"},"agent_actions":{"view_html":"https://pith.science/pith/SFCUMWGVZBMPBDTJLC5LLN3QXO","download_json":"https://pith.science/pith/SFCUMWGVZBMPBDTJLC5LLN3QXO.json","view_paper":"https://pith.science/paper/SFCUMWGV","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2507.03756&json=true","fetch_graph":"https://pith.science/api/pith-number/SFCUMWGVZBMPBDTJLC5LLN3QXO/graph.json","fetch_events":"https://pith.science/api/pith-number/SFCUMWGVZBMPBDTJLC5LLN3QXO/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/SFCUMWGVZBMPBDTJLC5LLN3QXO/action/timestamp_anchor","attest_storage":"https://pith.science/pith/SFCUMWGVZBMPBDTJLC5LLN3QXO/action/storage_attestation","attest_author":"https://pith.science/pith/SFCUMWGVZBMPBDTJLC5LLN3QXO/action/author_attestation","sign_citation":"https://pith.science/pith/SFCUMWGVZBMPBDTJLC5LLN3QXO/action/citation_signature","submit_replication":"https://pith.science/pith/SFCUMWGVZBMPBDTJLC5LLN3QXO/action/replication_record"}},"created_at":"2026-07-05T11:32:04.586268+00:00","updated_at":"2026-07-05T11:32:04.586268+00:00"}