{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:MWCJFWJATH7KYTILFJF2QN2DCM","short_pith_number":"pith:MWCJFWJA","schema_version":"1.0","canonical_sha256":"658492d92099feac4d0b2a4ba83743131e01868e04c55ff9e05727b949b582e2","source":{"kind":"arxiv","id":"2508.17689","version":1},"attestation_state":"computed","paper":{"title":"On the Edge of Memorization in Diffusion Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Druv Pai, Sam Buchanan, Valentin De Bortoli, Yi Ma","submitted_at":"2025-08-25T05:56:05Z","abstract_excerpt":"When do diffusion models reproduce their training data, and when are they able to generate samples beyond it? A practically relevant theoretical understanding of this interplay between memorization and generalization may significantly impact real-world deployments of diffusion models with respect to issues such as copyright infringement and data privacy. In this work, to disentangle the different factors that influence memorization and generalization in practical diffusion models, we introduce a scientific and mathematical \"laboratory\" for investigating these phenomena in diffusion models trai"},"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":"2508.17689","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-08-25T05:56:05Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"277d1905b08a44d5fed97601f13e5233bc9ad5e455c3ca68921a21d0f69f99f9","abstract_canon_sha256":"fbf9c1f222fdfcd8588a1a9a9f346e6de9fe090c0a4bf8de5a270d464d613f97"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:58:45.838395Z","signature_b64":"A+r5Gnfy8V2uPumJasn69tBqe+VTP9CmJBDK2O8JdeRWYwDETM2W+Xpl+wYeyMvMNjBp3D4mX+Yx8kF0OGMJAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"658492d92099feac4d0b2a4ba83743131e01868e04c55ff9e05727b949b582e2","last_reissued_at":"2026-07-05T11:58:45.837897Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:58:45.837897Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"On the Edge of Memorization in Diffusion Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Druv Pai, Sam Buchanan, Valentin De Bortoli, Yi Ma","submitted_at":"2025-08-25T05:56:05Z","abstract_excerpt":"When do diffusion models reproduce their training data, and when are they able to generate samples beyond it? A practically relevant theoretical understanding of this interplay between memorization and generalization may significantly impact real-world deployments of diffusion models with respect to issues such as copyright infringement and data privacy. In this work, to disentangle the different factors that influence memorization and generalization in practical diffusion models, we introduce a scientific and mathematical \"laboratory\" for investigating these phenomena in diffusion models trai"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2508.17689","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/2508.17689/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":"2508.17689","created_at":"2026-07-05T11:58:45.837956+00:00"},{"alias_kind":"arxiv_version","alias_value":"2508.17689v1","created_at":"2026-07-05T11:58:45.837956+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2508.17689","created_at":"2026-07-05T11:58:45.837956+00:00"},{"alias_kind":"pith_short_12","alias_value":"MWCJFWJATH7K","created_at":"2026-07-05T11:58:45.837956+00:00"},{"alias_kind":"pith_short_16","alias_value":"MWCJFWJATH7KYTIL","created_at":"2026-07-05T11:58:45.837956+00:00"},{"alias_kind":"pith_short_8","alias_value":"MWCJFWJA","created_at":"2026-07-05T11:58:45.837956+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":6,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.06624","citing_title":"Principles and Practice of Deep Representation Learning: or a Mathematical Theory of Memory","ref_index":11,"is_internal_anchor":false},{"citing_arxiv_id":"2603.13419","citing_title":"Diffusion Models Memorize in Training -- and Generalize in Inference","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2604.07213","citing_title":"Diffusion Processes on Implicit Manifolds","ref_index":15,"is_internal_anchor":false},{"citing_arxiv_id":"2605.19483","citing_title":"Adynamical systems view of training generativemodels and the memorization phenomenon","ref_index":20,"is_internal_anchor":false},{"citing_arxiv_id":"2604.07213","citing_title":"Diffusion Processes on Implicit Manifolds","ref_index":15,"is_internal_anchor":false},{"citing_arxiv_id":"2604.17673","citing_title":"Grokking of Diffusion Models: Case Study on Modular Addition","ref_index":2,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/MWCJFWJATH7KYTILFJF2QN2DCM","json":"https://pith.science/pith/MWCJFWJATH7KYTILFJF2QN2DCM.json","graph_json":"https://pith.science/api/pith-number/MWCJFWJATH7KYTILFJF2QN2DCM/graph.json","events_json":"https://pith.science/api/pith-number/MWCJFWJATH7KYTILFJF2QN2DCM/events.json","paper":"https://pith.science/paper/MWCJFWJA"},"agent_actions":{"view_html":"https://pith.science/pith/MWCJFWJATH7KYTILFJF2QN2DCM","download_json":"https://pith.science/pith/MWCJFWJATH7KYTILFJF2QN2DCM.json","view_paper":"https://pith.science/paper/MWCJFWJA","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2508.17689&json=true","fetch_graph":"https://pith.science/api/pith-number/MWCJFWJATH7KYTILFJF2QN2DCM/graph.json","fetch_events":"https://pith.science/api/pith-number/MWCJFWJATH7KYTILFJF2QN2DCM/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/MWCJFWJATH7KYTILFJF2QN2DCM/action/timestamp_anchor","attest_storage":"https://pith.science/pith/MWCJFWJATH7KYTILFJF2QN2DCM/action/storage_attestation","attest_author":"https://pith.science/pith/MWCJFWJATH7KYTILFJF2QN2DCM/action/author_attestation","sign_citation":"https://pith.science/pith/MWCJFWJATH7KYTILFJF2QN2DCM/action/citation_signature","submit_replication":"https://pith.science/pith/MWCJFWJATH7KYTILFJF2QN2DCM/action/replication_record"}},"created_at":"2026-07-05T11:58:45.837956+00:00","updated_at":"2026-07-05T11:58:45.837956+00:00"}