{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:XIENFPMMBD274PGSYR3G2ZAFCH","short_pith_number":"pith:XIENFPMM","schema_version":"1.0","canonical_sha256":"ba08d2bd8c08f5fe3cd2c4766d640511ebfdeb167d6ac520713b1382036ddf5e","source":{"kind":"arxiv","id":"2304.06027","version":2},"attestation_state":"computed","paper":{"title":"Continual Diffusion: Continual Customization of Text-to-Image Diffusion with C-LoRA","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CV","authors_text":"Hongxia Jin, James Seale Smith, Lingyu Zhang, Ting Hua, Yen-Chang Hsu, Yilin Shen, Zsolt Kira","submitted_at":"2023-04-12T17:59:41Z","abstract_excerpt":"Recent works demonstrate a remarkable ability to customize text-to-image diffusion models while only providing a few example images. What happens if you try to customize such models using multiple, fine-grained concepts in a sequential (i.e., continual) manner? In our work, we show that recent state-of-the-art customization of text-to-image models suffer from catastrophic forgetting when new concepts arrive sequentially. Specifically, when adding a new concept, the ability to generate high quality images of past, similar concepts degrade. To circumvent this forgetting, we propose a new method,"},"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":"2304.06027","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2023-04-12T17:59:41Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"b7ed3eca59b39830bfee0c58e6aba7b3688433a82dcc34bb64ebaf61790732e6","abstract_canon_sha256":"9dd49ad7cf6afb503eb0434716cd65b7d16cc81013a1f2c87342d82248c22c73"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:14:38.515199Z","signature_b64":"X2V9CzNiULuE2KoEk+yRaBU/Jm3MSNgcuLXU/wkkNkRTJE1K0jxlb6N6Vf8G3QDA6gZeRxQA6Q8llRxuQPIVCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ba08d2bd8c08f5fe3cd2c4766d640511ebfdeb167d6ac520713b1382036ddf5e","last_reissued_at":"2026-07-05T08:14:38.514756Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:14:38.514756Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Continual Diffusion: Continual Customization of Text-to-Image Diffusion with C-LoRA","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CV","authors_text":"Hongxia Jin, James Seale Smith, Lingyu Zhang, Ting Hua, Yen-Chang Hsu, Yilin Shen, Zsolt Kira","submitted_at":"2023-04-12T17:59:41Z","abstract_excerpt":"Recent works demonstrate a remarkable ability to customize text-to-image diffusion models while only providing a few example images. What happens if you try to customize such models using multiple, fine-grained concepts in a sequential (i.e., continual) manner? In our work, we show that recent state-of-the-art customization of text-to-image models suffer from catastrophic forgetting when new concepts arrive sequentially. Specifically, when adding a new concept, the ability to generate high quality images of past, similar concepts degrade. To circumvent this forgetting, we propose a new method,"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2304.06027","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/2304.06027/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":"2304.06027","created_at":"2026-07-05T08:14:38.514805+00:00"},{"alias_kind":"arxiv_version","alias_value":"2304.06027v2","created_at":"2026-07-05T08:14:38.514805+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2304.06027","created_at":"2026-07-05T08:14:38.514805+00:00"},{"alias_kind":"pith_short_12","alias_value":"XIENFPMMBD27","created_at":"2026-07-05T08:14:38.514805+00:00"},{"alias_kind":"pith_short_16","alias_value":"XIENFPMMBD274PGS","created_at":"2026-07-05T08:14:38.514805+00:00"},{"alias_kind":"pith_short_8","alias_value":"XIENFPMM","created_at":"2026-07-05T08:14:38.514805+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":7,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.09151","citing_title":"Customization under Fire: Plugin Poisoning in Text-to-Image Ecosystem","ref_index":73,"is_internal_anchor":false},{"citing_arxiv_id":"2605.20237","citing_title":"AnimeAdapter: A Modular Adapter for Appearance-Consistent Anime Character Generation","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2605.19750","citing_title":"CPC-VAR:Continual Personalized and Compositional Generation in Visual Autoregressive Models","ref_index":37,"is_internal_anchor":false},{"citing_arxiv_id":"2510.20512","citing_title":"Adversarial Concept Distillation for One-Step Diffusion Personalization","ref_index":77,"is_internal_anchor":false},{"citing_arxiv_id":"2603.20410","citing_title":"SLE-FNO: Single-Layer Extensions for Task-Agnostic Continual Learning in Fourier Neural Operators","ref_index":89,"is_internal_anchor":false},{"citing_arxiv_id":"2309.07864","citing_title":"The Rise and Potential of Large Language Model Based Agents: A Survey","ref_index":278,"is_internal_anchor":false},{"citing_arxiv_id":"2604.13730","citing_title":"ReConText3D: Replay-based Continual Text-to-3D Generation","ref_index":36,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/XIENFPMMBD274PGSYR3G2ZAFCH","json":"https://pith.science/pith/XIENFPMMBD274PGSYR3G2ZAFCH.json","graph_json":"https://pith.science/api/pith-number/XIENFPMMBD274PGSYR3G2ZAFCH/graph.json","events_json":"https://pith.science/api/pith-number/XIENFPMMBD274PGSYR3G2ZAFCH/events.json","paper":"https://pith.science/paper/XIENFPMM"},"agent_actions":{"view_html":"https://pith.science/pith/XIENFPMMBD274PGSYR3G2ZAFCH","download_json":"https://pith.science/pith/XIENFPMMBD274PGSYR3G2ZAFCH.json","view_paper":"https://pith.science/paper/XIENFPMM","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2304.06027&json=true","fetch_graph":"https://pith.science/api/pith-number/XIENFPMMBD274PGSYR3G2ZAFCH/graph.json","fetch_events":"https://pith.science/api/pith-number/XIENFPMMBD274PGSYR3G2ZAFCH/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/XIENFPMMBD274PGSYR3G2ZAFCH/action/timestamp_anchor","attest_storage":"https://pith.science/pith/XIENFPMMBD274PGSYR3G2ZAFCH/action/storage_attestation","attest_author":"https://pith.science/pith/XIENFPMMBD274PGSYR3G2ZAFCH/action/author_attestation","sign_citation":"https://pith.science/pith/XIENFPMMBD274PGSYR3G2ZAFCH/action/citation_signature","submit_replication":"https://pith.science/pith/XIENFPMMBD274PGSYR3G2ZAFCH/action/replication_record"}},"created_at":"2026-07-05T08:14:38.514805+00:00","updated_at":"2026-07-05T08:14:38.514805+00:00"}