{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:PZDBD4QHAEYXN6OIMKTXMUJ4KO","short_pith_number":"pith:PZDBD4QH","schema_version":"1.0","canonical_sha256":"7e4611f207013176f9c862a776513c53a6945a4fd03f4601b997a014827a2ced","source":{"kind":"arxiv","id":"2312.16145","version":2},"attestation_state":"computed","paper":{"title":"One-Dimensional Adapter to Rule Them All: Concepts, Diffusion Models and Erasing Applications","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CV","authors_text":"Guiguang Ding, Haiwen Hong, Hui Chen, Hui Xue, Jungong Han, Mengyao Lyu, Xuan Jin, Yuan He, Yuhong Yang","submitted_at":"2023-12-26T18:08:48Z","abstract_excerpt":"The prevalent use of commercial and open-source diffusion models (DMs) for text-to-image generation prompts risk mitigation to prevent undesired behaviors. Existing concept erasing methods in academia are all based on full parameter or specification-based fine-tuning, from which we observe the following issues: 1) Generation alternation towards erosion: Parameter drift during target elimination causes alternations and potential deformations across all generations, even eroding other concepts at varying degrees, which is more evident with multi-concept erased; 2) Transfer inability & deployment"},"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":"2312.16145","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2023-12-26T18:08:48Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"8ee2bd1b16fcd38adb7b964cb25f0df7a2b63832d8e1c809be0dbeb304a183ed","abstract_canon_sha256":"af0f468ba5df8f656aff8a21be5600a93efbda2ddc5df3060f88f841419f13c3"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:54:54.020748Z","signature_b64":"cAbHe+hJPc3vauSjtBxDjSzpYme6Y8+UpQJKKnm7leWVMCtKu/bYIq3pAGSo0PStjsVxZt+acGjQoA1RDoMMCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7e4611f207013176f9c862a776513c53a6945a4fd03f4601b997a014827a2ced","last_reissued_at":"2026-07-05T07:54:54.020331Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:54:54.020331Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"One-Dimensional Adapter to Rule Them All: Concepts, Diffusion Models and Erasing Applications","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CV","authors_text":"Guiguang Ding, Haiwen Hong, Hui Chen, Hui Xue, Jungong Han, Mengyao Lyu, Xuan Jin, Yuan He, Yuhong Yang","submitted_at":"2023-12-26T18:08:48Z","abstract_excerpt":"The prevalent use of commercial and open-source diffusion models (DMs) for text-to-image generation prompts risk mitigation to prevent undesired behaviors. Existing concept erasing methods in academia are all based on full parameter or specification-based fine-tuning, from which we observe the following issues: 1) Generation alternation towards erosion: Parameter drift during target elimination causes alternations and potential deformations across all generations, even eroding other concepts at varying degrees, which is more evident with multi-concept erased; 2) Transfer inability & deployment"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2312.16145","kind":"arxiv","version":2},"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/2312.16145/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":"2312.16145","created_at":"2026-07-05T07:54:54.020395+00:00"},{"alias_kind":"arxiv_version","alias_value":"2312.16145v2","created_at":"2026-07-05T07:54:54.020395+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2312.16145","created_at":"2026-07-05T07:54:54.020395+00:00"},{"alias_kind":"pith_short_12","alias_value":"PZDBD4QHAEYX","created_at":"2026-07-05T07:54:54.020395+00:00"},{"alias_kind":"pith_short_16","alias_value":"PZDBD4QHAEYXN6OI","created_at":"2026-07-05T07:54:54.020395+00:00"},{"alias_kind":"pith_short_8","alias_value":"PZDBD4QH","created_at":"2026-07-05T07:54:54.020395+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.01658","citing_title":"CoreUnlearn: Rethinking Concept Unlearning through Disentangled Component-Level Erasure in Text-guided Diffusion Models","ref_index":33,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/PZDBD4QHAEYXN6OIMKTXMUJ4KO","json":"https://pith.science/pith/PZDBD4QHAEYXN6OIMKTXMUJ4KO.json","graph_json":"https://pith.science/api/pith-number/PZDBD4QHAEYXN6OIMKTXMUJ4KO/graph.json","events_json":"https://pith.science/api/pith-number/PZDBD4QHAEYXN6OIMKTXMUJ4KO/events.json","paper":"https://pith.science/paper/PZDBD4QH"},"agent_actions":{"view_html":"https://pith.science/pith/PZDBD4QHAEYXN6OIMKTXMUJ4KO","download_json":"https://pith.science/pith/PZDBD4QHAEYXN6OIMKTXMUJ4KO.json","view_paper":"https://pith.science/paper/PZDBD4QH","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2312.16145&json=true","fetch_graph":"https://pith.science/api/pith-number/PZDBD4QHAEYXN6OIMKTXMUJ4KO/graph.json","fetch_events":"https://pith.science/api/pith-number/PZDBD4QHAEYXN6OIMKTXMUJ4KO/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/PZDBD4QHAEYXN6OIMKTXMUJ4KO/action/timestamp_anchor","attest_storage":"https://pith.science/pith/PZDBD4QHAEYXN6OIMKTXMUJ4KO/action/storage_attestation","attest_author":"https://pith.science/pith/PZDBD4QHAEYXN6OIMKTXMUJ4KO/action/author_attestation","sign_citation":"https://pith.science/pith/PZDBD4QHAEYXN6OIMKTXMUJ4KO/action/citation_signature","submit_replication":"https://pith.science/pith/PZDBD4QHAEYXN6OIMKTXMUJ4KO/action/replication_record"}},"created_at":"2026-07-05T07:54:54.020395+00:00","updated_at":"2026-07-05T07:54:54.020395+00:00"}