{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:2JMSNKIU4DLQUOLJ73MG7EZU3H","short_pith_number":"pith:2JMSNKIU","schema_version":"1.0","canonical_sha256":"d25926a914e0d70a3969fed86f9334d9e41b296ab84e3da7d19a0a30cb318031","source":{"kind":"arxiv","id":"2508.12148","version":1},"attestation_state":"computed","paper":{"title":"Demystifying Foreground-Background Memorization in Diffusion Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Adam Dziedzic, Franziska Boenisch, Gautam Kamath, Jimmy Z. Di, Yaoliang Yu, Yiwei Lu","submitted_at":"2025-08-16T20:15:16Z","abstract_excerpt":"Diffusion models (DMs) memorize training images and can reproduce near-duplicates during generation. Current detection methods identify verbatim memorization but fail to capture two critical aspects: quantifying partial memorization occurring in small image regions, and memorization patterns beyond specific prompt-image pairs. To address these limitations, we propose Foreground Background Memorization (FB-Mem), a novel segmentation-based metric that classifies and quantifies memorized regions within generated images. Our method reveals that memorization is more pervasive than previously unders"},"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":"2508.12148","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2025-08-16T20:15:16Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"37f14e688db755447643d566a41d52596f07f3d2310cce0d258c97600242ad8d","abstract_canon_sha256":"e078e6ea73b321feda0f6c06eb7bf0a0de6aceba8988dc5cb62dcb107bb5e70c"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:55:00.907380Z","signature_b64":"i167QcXb8UGYvkiGX8zhBhuvz8lHz9F9mV9jmsVROrHmjAjzdmIF4HZ8sF28OyHIvooIikkqFKbJIxtUP5lACQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d25926a914e0d70a3969fed86f9334d9e41b296ab84e3da7d19a0a30cb318031","last_reissued_at":"2026-07-05T11:55:00.906936Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:55:00.906936Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Demystifying Foreground-Background Memorization in Diffusion Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Adam Dziedzic, Franziska Boenisch, Gautam Kamath, Jimmy Z. Di, Yaoliang Yu, Yiwei Lu","submitted_at":"2025-08-16T20:15:16Z","abstract_excerpt":"Diffusion models (DMs) memorize training images and can reproduce near-duplicates during generation. Current detection methods identify verbatim memorization but fail to capture two critical aspects: quantifying partial memorization occurring in small image regions, and memorization patterns beyond specific prompt-image pairs. To address these limitations, we propose Foreground Background Memorization (FB-Mem), a novel segmentation-based metric that classifies and quantifies memorized regions within generated images. Our method reveals that memorization is more pervasive than previously unders"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2508.12148","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/2508.12148/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":"2508.12148","created_at":"2026-07-05T11:55:00.906991+00:00"},{"alias_kind":"arxiv_version","alias_value":"2508.12148v1","created_at":"2026-07-05T11:55:00.906991+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2508.12148","created_at":"2026-07-05T11:55:00.906991+00:00"},{"alias_kind":"pith_short_12","alias_value":"2JMSNKIU4DLQ","created_at":"2026-07-05T11:55:00.906991+00:00"},{"alias_kind":"pith_short_16","alias_value":"2JMSNKIU4DLQUOLJ","created_at":"2026-07-05T11:55:00.906991+00:00"},{"alias_kind":"pith_short_8","alias_value":"2JMSNKIU","created_at":"2026-07-05T11:55:00.906991+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":15,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/2JMSNKIU4DLQUOLJ73MG7EZU3H","json":"https://pith.science/pith/2JMSNKIU4DLQUOLJ73MG7EZU3H.json","graph_json":"https://pith.science/api/pith-number/2JMSNKIU4DLQUOLJ73MG7EZU3H/graph.json","events_json":"https://pith.science/api/pith-number/2JMSNKIU4DLQUOLJ73MG7EZU3H/events.json","paper":"https://pith.science/paper/2JMSNKIU"},"agent_actions":{"view_html":"https://pith.science/pith/2JMSNKIU4DLQUOLJ73MG7EZU3H","download_json":"https://pith.science/pith/2JMSNKIU4DLQUOLJ73MG7EZU3H.json","view_paper":"https://pith.science/paper/2JMSNKIU","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2508.12148&json=true","fetch_graph":"https://pith.science/api/pith-number/2JMSNKIU4DLQUOLJ73MG7EZU3H/graph.json","fetch_events":"https://pith.science/api/pith-number/2JMSNKIU4DLQUOLJ73MG7EZU3H/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/2JMSNKIU4DLQUOLJ73MG7EZU3H/action/timestamp_anchor","attest_storage":"https://pith.science/pith/2JMSNKIU4DLQUOLJ73MG7EZU3H/action/storage_attestation","attest_author":"https://pith.science/pith/2JMSNKIU4DLQUOLJ73MG7EZU3H/action/author_attestation","sign_citation":"https://pith.science/pith/2JMSNKIU4DLQUOLJ73MG7EZU3H/action/citation_signature","submit_replication":"https://pith.science/pith/2JMSNKIU4DLQUOLJ73MG7EZU3H/action/replication_record"}},"created_at":"2026-07-05T11:55:00.906991+00:00","updated_at":"2026-07-05T11:55:00.906991+00:00"}