{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:S6H7FBNQY3CLM26KSUHZZDIG3M","short_pith_number":"pith:S6H7FBNQ","schema_version":"1.0","canonical_sha256":"978ff285b0c6c4b66bca950f9c8d06db0897289264da3ff73ffea1e3956875e1","source":{"kind":"arxiv","id":"2506.18428","version":1},"attestation_state":"computed","paper":{"title":"How Robust is Model Editing after Fine-Tuning? An Empirical Study on Text-to-Image Diffusion Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.AI","authors_text":"Feng He, Marco Valentino, Zhenyang Liu, Zhixue Zhao","submitted_at":"2025-06-23T09:10:29Z","abstract_excerpt":"Model editing offers a low-cost technique to inject or correct a particular behavior in a pre-trained model without extensive retraining, supporting applications such as factual correction and bias mitigation. Despite this common practice, it remains unknown whether edits persist after fine-tuning or whether they are inadvertently reversed. This question has fundamental practical implications. For example, if fine-tuning removes prior edits, it could serve as a defence mechanism against hidden malicious edits. Vice versa, the unintended removal of edits related to bias mitigation could pose se"},"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":"2506.18428","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2025-06-23T09:10:29Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"eb97214d3c95ef530222fe02b7e4f5c56c549c05cef6470204f13a953a5c9b16","abstract_canon_sha256":"9b8ff070c2f435311d80547bd42b44cd91d3ac953602b3f0b396fa09c1186288"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:25:50.127722Z","signature_b64":"+zxvlhxmga+Sh8xopgOB1mcKbmcQPIDNDUbAvUmpaPNz45HLujIC/ZM338JvNLHUax4DidXL/CH2CAwmuvtjDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"978ff285b0c6c4b66bca950f9c8d06db0897289264da3ff73ffea1e3956875e1","last_reissued_at":"2026-07-05T11:25:50.127251Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:25:50.127251Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"How Robust is Model Editing after Fine-Tuning? An Empirical Study on Text-to-Image Diffusion Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.AI","authors_text":"Feng He, Marco Valentino, Zhenyang Liu, Zhixue Zhao","submitted_at":"2025-06-23T09:10:29Z","abstract_excerpt":"Model editing offers a low-cost technique to inject or correct a particular behavior in a pre-trained model without extensive retraining, supporting applications such as factual correction and bias mitigation. Despite this common practice, it remains unknown whether edits persist after fine-tuning or whether they are inadvertently reversed. This question has fundamental practical implications. For example, if fine-tuning removes prior edits, it could serve as a defence mechanism against hidden malicious edits. Vice versa, the unintended removal of edits related to bias mitigation could pose se"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.18428","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/2506.18428/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":"2506.18428","created_at":"2026-07-05T11:25:50.127310+00:00"},{"alias_kind":"arxiv_version","alias_value":"2506.18428v1","created_at":"2026-07-05T11:25:50.127310+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.18428","created_at":"2026-07-05T11:25:50.127310+00:00"},{"alias_kind":"pith_short_12","alias_value":"S6H7FBNQY3CL","created_at":"2026-07-05T11:25:50.127310+00:00"},{"alias_kind":"pith_short_16","alias_value":"S6H7FBNQY3CLM26K","created_at":"2026-07-05T11:25:50.127310+00:00"},{"alias_kind":"pith_short_8","alias_value":"S6H7FBNQ","created_at":"2026-07-05T11:25:50.127310+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/S6H7FBNQY3CLM26KSUHZZDIG3M","json":"https://pith.science/pith/S6H7FBNQY3CLM26KSUHZZDIG3M.json","graph_json":"https://pith.science/api/pith-number/S6H7FBNQY3CLM26KSUHZZDIG3M/graph.json","events_json":"https://pith.science/api/pith-number/S6H7FBNQY3CLM26KSUHZZDIG3M/events.json","paper":"https://pith.science/paper/S6H7FBNQ"},"agent_actions":{"view_html":"https://pith.science/pith/S6H7FBNQY3CLM26KSUHZZDIG3M","download_json":"https://pith.science/pith/S6H7FBNQY3CLM26KSUHZZDIG3M.json","view_paper":"https://pith.science/paper/S6H7FBNQ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2506.18428&json=true","fetch_graph":"https://pith.science/api/pith-number/S6H7FBNQY3CLM26KSUHZZDIG3M/graph.json","fetch_events":"https://pith.science/api/pith-number/S6H7FBNQY3CLM26KSUHZZDIG3M/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/S6H7FBNQY3CLM26KSUHZZDIG3M/action/timestamp_anchor","attest_storage":"https://pith.science/pith/S6H7FBNQY3CLM26KSUHZZDIG3M/action/storage_attestation","attest_author":"https://pith.science/pith/S6H7FBNQY3CLM26KSUHZZDIG3M/action/author_attestation","sign_citation":"https://pith.science/pith/S6H7FBNQY3CLM26KSUHZZDIG3M/action/citation_signature","submit_replication":"https://pith.science/pith/S6H7FBNQY3CLM26KSUHZZDIG3M/action/replication_record"}},"created_at":"2026-07-05T11:25:50.127310+00:00","updated_at":"2026-07-05T11:25:50.127310+00:00"}