{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:2DSMEAVWT463KBTYLXGEQCFL7P","short_pith_number":"pith:2DSMEAVW","schema_version":"1.0","canonical_sha256":"d0e4c202b69f3db506785dcc4808abfbf832828ee85463175b33d6f457987516","source":{"kind":"arxiv","id":"2402.05947","version":1},"attestation_state":"computed","paper":{"title":"Separable Multi-Concept Erasure from Diffusion Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV"],"primary_cat":"cs.LG","authors_text":"Baocai Yin, Lihe Zhang, Mengnan Zhao, Tianhang Zheng, Yuqiu Kong","submitted_at":"2024-02-03T11:10:57Z","abstract_excerpt":"Large-scale diffusion models, known for their impressive image generation capabilities, have raised concerns among researchers regarding social impacts, such as the imitation of copyrighted artistic styles. In response, existing approaches turn to machine unlearning techniques to eliminate unsafe concepts from pre-trained models. However, these methods compromise the generative performance and neglect the coupling among multi-concept erasures, as well as the concept restoration problem. To address these issues, we propose a Separable Multi-concept Eraser (SepME), which mainly includes two part"},"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":"2402.05947","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-02-03T11:10:57Z","cross_cats_sorted":["cs.CV"],"title_canon_sha256":"163832536acf982c480a76dc6359d9e1de464813e098528b8e38fb19b50441d3","abstract_canon_sha256":"c75af089f1ce9e67566b5a9e054d140ef4aed583f7cf026f1e4655de45175163"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:43:15.788520Z","signature_b64":"fiOH0f5MtoDy6NW5IaiQNUHsmh59oC0HsZzzyrgXcoew1YLtNw12bgx/X5V4he5rurkBn5Q4H+XOmgCMEIOVBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d0e4c202b69f3db506785dcc4808abfbf832828ee85463175b33d6f457987516","last_reissued_at":"2026-07-05T07:43:15.787961Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:43:15.787961Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Separable Multi-Concept Erasure from Diffusion Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV"],"primary_cat":"cs.LG","authors_text":"Baocai Yin, Lihe Zhang, Mengnan Zhao, Tianhang Zheng, Yuqiu Kong","submitted_at":"2024-02-03T11:10:57Z","abstract_excerpt":"Large-scale diffusion models, known for their impressive image generation capabilities, have raised concerns among researchers regarding social impacts, such as the imitation of copyrighted artistic styles. In response, existing approaches turn to machine unlearning techniques to eliminate unsafe concepts from pre-trained models. However, these methods compromise the generative performance and neglect the coupling among multi-concept erasures, as well as the concept restoration problem. To address these issues, we propose a Separable Multi-concept Eraser (SepME), which mainly includes two part"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2402.05947","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/2402.05947/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":"2402.05947","created_at":"2026-07-05T07:43:15.788029+00:00"},{"alias_kind":"arxiv_version","alias_value":"2402.05947v1","created_at":"2026-07-05T07:43:15.788029+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2402.05947","created_at":"2026-07-05T07:43:15.788029+00:00"},{"alias_kind":"pith_short_12","alias_value":"2DSMEAVWT463","created_at":"2026-07-05T07:43:15.788029+00:00"},{"alias_kind":"pith_short_16","alias_value":"2DSMEAVWT463KBTY","created_at":"2026-07-05T07:43:15.788029+00:00"},{"alias_kind":"pith_short_8","alias_value":"2DSMEAVW","created_at":"2026-07-05T07:43:15.788029+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":4,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2607.07907","citing_title":"Multimodal Unlearning Across Vision, Language, Video, and Audio: Survey of Methods, Datasets, and Benchmarks","ref_index":86,"is_internal_anchor":true},{"citing_arxiv_id":"2606.01658","citing_title":"CoreUnlearn: Rethinking Concept Unlearning through Disentangled Component-Level Erasure in Text-guided Diffusion Models","ref_index":67,"is_internal_anchor":false},{"citing_arxiv_id":"2601.06163","citing_title":"Forget-It-All: Multi-Concept Machine Unlearning via Concept-Aware Neuron Masking","ref_index":14,"is_internal_anchor":false},{"citing_arxiv_id":"2604.10032","citing_title":"Closed-Form Concept Erasure via Double Projections","ref_index":74,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/2DSMEAVWT463KBTYLXGEQCFL7P","json":"https://pith.science/pith/2DSMEAVWT463KBTYLXGEQCFL7P.json","graph_json":"https://pith.science/api/pith-number/2DSMEAVWT463KBTYLXGEQCFL7P/graph.json","events_json":"https://pith.science/api/pith-number/2DSMEAVWT463KBTYLXGEQCFL7P/events.json","paper":"https://pith.science/paper/2DSMEAVW"},"agent_actions":{"view_html":"https://pith.science/pith/2DSMEAVWT463KBTYLXGEQCFL7P","download_json":"https://pith.science/pith/2DSMEAVWT463KBTYLXGEQCFL7P.json","view_paper":"https://pith.science/paper/2DSMEAVW","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2402.05947&json=true","fetch_graph":"https://pith.science/api/pith-number/2DSMEAVWT463KBTYLXGEQCFL7P/graph.json","fetch_events":"https://pith.science/api/pith-number/2DSMEAVWT463KBTYLXGEQCFL7P/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/2DSMEAVWT463KBTYLXGEQCFL7P/action/timestamp_anchor","attest_storage":"https://pith.science/pith/2DSMEAVWT463KBTYLXGEQCFL7P/action/storage_attestation","attest_author":"https://pith.science/pith/2DSMEAVWT463KBTYLXGEQCFL7P/action/author_attestation","sign_citation":"https://pith.science/pith/2DSMEAVWT463KBTYLXGEQCFL7P/action/citation_signature","submit_replication":"https://pith.science/pith/2DSMEAVWT463KBTYLXGEQCFL7P/action/replication_record"}},"created_at":"2026-07-05T07:43:15.788029+00:00","updated_at":"2026-07-05T07:43:15.788029+00:00"}