{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:UBKX3ZTRTLAVHMKIWTRXFMAL3H","short_pith_number":"pith:UBKX3ZTR","schema_version":"1.0","canonical_sha256":"a0557de6719ac153b148b4e372b00bd9c175c707482365721f5b1cf38a3b5b58","source":{"kind":"arxiv","id":"2509.00488","version":1},"attestation_state":"computed","paper":{"title":"Localizing and Mitigating Memorization in Image Autoregressive Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Adam Dziedzic, Aditya Kasliwal, Franziska Boenisch","submitted_at":"2025-08-30T13:04:36Z","abstract_excerpt":"Image AutoRegressive (IAR) models have achieved state-of-the-art performance in speed and quality of generated images. However, they also raise concerns about memorization of their training data and its implications for privacy. This work explores where and how such memorization occurs within different image autoregressive architectures by measuring a fine-grained memorization. The analysis reveals that memorization patterns differ across various architectures of IARs. In hierarchical per-resolution architectures, it tends to emerge early and deepen with resolutions, while in IARs with standar"},"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":"2509.00488","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-08-30T13:04:36Z","cross_cats_sorted":[],"title_canon_sha256":"405189733fd477f8696331d491aeb881af2aaa107af489bd340d4a7532870bfe","abstract_canon_sha256":"16dbc886649f772e1ddb9bec8250749f8072473910a0fceb64de8a06787b80a3"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T12:02:17.107468Z","signature_b64":"ol+XW3trl8ZSTVdg4NBIkGKZuWyPgNNoisjB06+WHllDvd6C01NSmHihl7kFDI4KzfxMFMjnq+veDzT2/Tw0CQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a0557de6719ac153b148b4e372b00bd9c175c707482365721f5b1cf38a3b5b58","last_reissued_at":"2026-07-05T12:02:17.106794Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T12:02:17.106794Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Localizing and Mitigating Memorization in Image Autoregressive Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Adam Dziedzic, Aditya Kasliwal, Franziska Boenisch","submitted_at":"2025-08-30T13:04:36Z","abstract_excerpt":"Image AutoRegressive (IAR) models have achieved state-of-the-art performance in speed and quality of generated images. However, they also raise concerns about memorization of their training data and its implications for privacy. This work explores where and how such memorization occurs within different image autoregressive architectures by measuring a fine-grained memorization. The analysis reveals that memorization patterns differ across various architectures of IARs. In hierarchical per-resolution architectures, it tends to emerge early and deepen with resolutions, while in IARs with standar"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2509.00488","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/2509.00488/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":"2509.00488","created_at":"2026-07-05T12:02:17.106872+00:00"},{"alias_kind":"arxiv_version","alias_value":"2509.00488v1","created_at":"2026-07-05T12:02:17.106872+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2509.00488","created_at":"2026-07-05T12:02:17.106872+00:00"},{"alias_kind":"pith_short_12","alias_value":"UBKX3ZTRTLAV","created_at":"2026-07-05T12:02:17.106872+00:00"},{"alias_kind":"pith_short_16","alias_value":"UBKX3ZTRTLAVHMKI","created_at":"2026-07-05T12:02:17.106872+00:00"},{"alias_kind":"pith_short_8","alias_value":"UBKX3ZTR","created_at":"2026-07-05T12:02:17.106872+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/UBKX3ZTRTLAVHMKIWTRXFMAL3H","json":"https://pith.science/pith/UBKX3ZTRTLAVHMKIWTRXFMAL3H.json","graph_json":"https://pith.science/api/pith-number/UBKX3ZTRTLAVHMKIWTRXFMAL3H/graph.json","events_json":"https://pith.science/api/pith-number/UBKX3ZTRTLAVHMKIWTRXFMAL3H/events.json","paper":"https://pith.science/paper/UBKX3ZTR"},"agent_actions":{"view_html":"https://pith.science/pith/UBKX3ZTRTLAVHMKIWTRXFMAL3H","download_json":"https://pith.science/pith/UBKX3ZTRTLAVHMKIWTRXFMAL3H.json","view_paper":"https://pith.science/paper/UBKX3ZTR","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2509.00488&json=true","fetch_graph":"https://pith.science/api/pith-number/UBKX3ZTRTLAVHMKIWTRXFMAL3H/graph.json","fetch_events":"https://pith.science/api/pith-number/UBKX3ZTRTLAVHMKIWTRXFMAL3H/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/UBKX3ZTRTLAVHMKIWTRXFMAL3H/action/timestamp_anchor","attest_storage":"https://pith.science/pith/UBKX3ZTRTLAVHMKIWTRXFMAL3H/action/storage_attestation","attest_author":"https://pith.science/pith/UBKX3ZTRTLAVHMKIWTRXFMAL3H/action/author_attestation","sign_citation":"https://pith.science/pith/UBKX3ZTRTLAVHMKIWTRXFMAL3H/action/citation_signature","submit_replication":"https://pith.science/pith/UBKX3ZTRTLAVHMKIWTRXFMAL3H/action/replication_record"}},"created_at":"2026-07-05T12:02:17.106872+00:00","updated_at":"2026-07-05T12:02:17.106872+00:00"}