{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:HYJ6A6GDJTYEG2X6KSPFNZYYQQ","short_pith_number":"pith:HYJ6A6GD","schema_version":"1.0","canonical_sha256":"3e13e078c34cf0436afe549e56e718842e25eebb33be622fff7558343a4b84df","source":{"kind":"arxiv","id":"2508.08812","version":1},"attestation_state":"computed","paper":{"title":"TARA: Token-Aware LoRA for Composable Personalization in Diffusion Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Jianzhuang Liu, Lingtao Zheng, Mingfu Yan, Shifeng Chen, Yi Huang, Yufeng Yang, Yuqi Peng","submitted_at":"2025-08-12T10:14:15Z","abstract_excerpt":"Personalized text-to-image generation aims to synthesize novel images of a specific subject or style using only a few reference images. Recent methods based on Low-Rank Adaptation (LoRA) enable efficient single-concept customization by injecting lightweight, concept-specific adapters into pre-trained diffusion models. However, combining multiple LoRA modules for multi-concept generation often leads to identity missing and visual feature leakage. In this work, we identify two key issues behind these failures: (1) token-wise interference among different LoRA modules, and (2) spatial misalignment"},"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":"2508.08812","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2025-08-12T10:14:15Z","cross_cats_sorted":[],"title_canon_sha256":"a1a5b0dcaa73ca799584ef7cfe3ed86293c0d9a9e2b2a1eafe593760a1985cde","abstract_canon_sha256":"096cd84782ddce3378399d4a131feeb42f7e7efa544879d18d6ef4babc26363d"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:52:34.667234Z","signature_b64":"6h+iAa40U+eE8xnmqJuQfPBU+1wugfnnwA3au+QzWXDKgR8JsnINrNlyzyBOoB+C0eebGjWzB3gugJrKqOTIBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3e13e078c34cf0436afe549e56e718842e25eebb33be622fff7558343a4b84df","last_reissued_at":"2026-07-05T11:52:34.666743Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:52:34.666743Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"TARA: Token-Aware LoRA for Composable Personalization in Diffusion Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Jianzhuang Liu, Lingtao Zheng, Mingfu Yan, Shifeng Chen, Yi Huang, Yufeng Yang, Yuqi Peng","submitted_at":"2025-08-12T10:14:15Z","abstract_excerpt":"Personalized text-to-image generation aims to synthesize novel images of a specific subject or style using only a few reference images. Recent methods based on Low-Rank Adaptation (LoRA) enable efficient single-concept customization by injecting lightweight, concept-specific adapters into pre-trained diffusion models. However, combining multiple LoRA modules for multi-concept generation often leads to identity missing and visual feature leakage. In this work, we identify two key issues behind these failures: (1) token-wise interference among different LoRA modules, and (2) spatial misalignment"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2508.08812","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/2508.08812/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":"2508.08812","created_at":"2026-07-05T11:52:34.666803+00:00"},{"alias_kind":"arxiv_version","alias_value":"2508.08812v1","created_at":"2026-07-05T11:52:34.666803+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2508.08812","created_at":"2026-07-05T11:52:34.666803+00:00"},{"alias_kind":"pith_short_12","alias_value":"HYJ6A6GDJTYE","created_at":"2026-07-05T11:52:34.666803+00:00"},{"alias_kind":"pith_short_16","alias_value":"HYJ6A6GDJTYEG2X6","created_at":"2026-07-05T11:52:34.666803+00:00"},{"alias_kind":"pith_short_8","alias_value":"HYJ6A6GD","created_at":"2026-07-05T11:52:34.666803+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.03792","citing_title":"Training-Free Multi-Concept LoRA Composition with Prompt-Aware Weighting","ref_index":40,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/HYJ6A6GDJTYEG2X6KSPFNZYYQQ","json":"https://pith.science/pith/HYJ6A6GDJTYEG2X6KSPFNZYYQQ.json","graph_json":"https://pith.science/api/pith-number/HYJ6A6GDJTYEG2X6KSPFNZYYQQ/graph.json","events_json":"https://pith.science/api/pith-number/HYJ6A6GDJTYEG2X6KSPFNZYYQQ/events.json","paper":"https://pith.science/paper/HYJ6A6GD"},"agent_actions":{"view_html":"https://pith.science/pith/HYJ6A6GDJTYEG2X6KSPFNZYYQQ","download_json":"https://pith.science/pith/HYJ6A6GDJTYEG2X6KSPFNZYYQQ.json","view_paper":"https://pith.science/paper/HYJ6A6GD","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2508.08812&json=true","fetch_graph":"https://pith.science/api/pith-number/HYJ6A6GDJTYEG2X6KSPFNZYYQQ/graph.json","fetch_events":"https://pith.science/api/pith-number/HYJ6A6GDJTYEG2X6KSPFNZYYQQ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/HYJ6A6GDJTYEG2X6KSPFNZYYQQ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/HYJ6A6GDJTYEG2X6KSPFNZYYQQ/action/storage_attestation","attest_author":"https://pith.science/pith/HYJ6A6GDJTYEG2X6KSPFNZYYQQ/action/author_attestation","sign_citation":"https://pith.science/pith/HYJ6A6GDJTYEG2X6KSPFNZYYQQ/action/citation_signature","submit_replication":"https://pith.science/pith/HYJ6A6GDJTYEG2X6KSPFNZYYQQ/action/replication_record"}},"created_at":"2026-07-05T11:52:34.666803+00:00","updated_at":"2026-07-05T11:52:34.666803+00:00"}