{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:YLEWLVDBJ7OXNWEYWLZABJF3S5","short_pith_number":"pith:YLEWLVDB","schema_version":"1.0","canonical_sha256":"c2c965d4614fdd76d898b2f200a4bb9765c2fe676803f2b7966bb396b6d2718f","source":{"kind":"arxiv","id":"2412.05148","version":2},"attestation_state":"computed","paper":{"title":"LoRA.rar: Learning to Merge LoRAs via Hypernetworks for Subject-Style Conditioned Image Generation","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CV","authors_text":"Donald Shenaj, Mete Ozay, Ondrej Bohdal, Pietro Zanuttigh, Umberto Michieli","submitted_at":"2024-12-06T16:04:56Z","abstract_excerpt":"Recent advancements in image generation models have enabled personalized image creation with both user-defined subjects (content) and styles. Prior works achieved personalization by merging corresponding low-rank adapters (LoRAs) through optimization-based methods, which are computationally demanding and unsuitable for real-time use on resource-constrained devices like smartphones. To address this, we introduce LoRA$.$rar, a method that not only improves image quality but also achieves a remarkable speedup of over $4000\\times$ in the merging process. We collect a dataset of style and subject L"},"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":"2412.05148","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CV","submitted_at":"2024-12-06T16:04:56Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"040be3fdb225b3ae42676067b107a2603000f17ab5e61a70d9882748e43f72d0","abstract_canon_sha256":"a32c76cd11b4e4dc119569911c6be796d54d0ee20af2130dc013f7a7391026d6"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:51:19.744164Z","signature_b64":"/HxJQdLvXqrbHi/CaKFG/BZtP+OlYL5y1Cr077KKUr9jXZo88m8/utw/YFmU8sCsPV0e7WNsyTZ3CFlk2uzvCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c2c965d4614fdd76d898b2f200a4bb9765c2fe676803f2b7966bb396b6d2718f","last_reissued_at":"2026-07-05T11:51:19.743648Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:51:19.743648Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"LoRA.rar: Learning to Merge LoRAs via Hypernetworks for Subject-Style Conditioned Image Generation","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CV","authors_text":"Donald Shenaj, Mete Ozay, Ondrej Bohdal, Pietro Zanuttigh, Umberto Michieli","submitted_at":"2024-12-06T16:04:56Z","abstract_excerpt":"Recent advancements in image generation models have enabled personalized image creation with both user-defined subjects (content) and styles. Prior works achieved personalization by merging corresponding low-rank adapters (LoRAs) through optimization-based methods, which are computationally demanding and unsuitable for real-time use on resource-constrained devices like smartphones. To address this, we introduce LoRA$.$rar, a method that not only improves image quality but also achieves a remarkable speedup of over $4000\\times$ in the merging process. We collect a dataset of style and subject L"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.05148","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/2412.05148/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":"2412.05148","created_at":"2026-07-05T11:51:19.743716+00:00"},{"alias_kind":"arxiv_version","alias_value":"2412.05148v2","created_at":"2026-07-05T11:51:19.743716+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.05148","created_at":"2026-07-05T11:51:19.743716+00:00"},{"alias_kind":"pith_short_12","alias_value":"YLEWLVDBJ7OX","created_at":"2026-07-05T11:51:19.743716+00:00"},{"alias_kind":"pith_short_16","alias_value":"YLEWLVDBJ7OXNWEY","created_at":"2026-07-05T11:51:19.743716+00:00"},{"alias_kind":"pith_short_8","alias_value":"YLEWLVDB","created_at":"2026-07-05T11:51:19.743716+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2604.23750","citing_title":"The Override Gap: A Magnitude Account of Knowledge Conflict Failure in Hypernetwork-Based Instant LLM Adaptation","ref_index":31,"is_internal_anchor":false},{"citing_arxiv_id":"2604.23750","citing_title":"The Override Gap: A Magnitude Account of Knowledge Conflict Failure in Hypernetwork-Based Instant LLM Adaptation","ref_index":31,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/YLEWLVDBJ7OXNWEYWLZABJF3S5","json":"https://pith.science/pith/YLEWLVDBJ7OXNWEYWLZABJF3S5.json","graph_json":"https://pith.science/api/pith-number/YLEWLVDBJ7OXNWEYWLZABJF3S5/graph.json","events_json":"https://pith.science/api/pith-number/YLEWLVDBJ7OXNWEYWLZABJF3S5/events.json","paper":"https://pith.science/paper/YLEWLVDB"},"agent_actions":{"view_html":"https://pith.science/pith/YLEWLVDBJ7OXNWEYWLZABJF3S5","download_json":"https://pith.science/pith/YLEWLVDBJ7OXNWEYWLZABJF3S5.json","view_paper":"https://pith.science/paper/YLEWLVDB","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2412.05148&json=true","fetch_graph":"https://pith.science/api/pith-number/YLEWLVDBJ7OXNWEYWLZABJF3S5/graph.json","fetch_events":"https://pith.science/api/pith-number/YLEWLVDBJ7OXNWEYWLZABJF3S5/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/YLEWLVDBJ7OXNWEYWLZABJF3S5/action/timestamp_anchor","attest_storage":"https://pith.science/pith/YLEWLVDBJ7OXNWEYWLZABJF3S5/action/storage_attestation","attest_author":"https://pith.science/pith/YLEWLVDBJ7OXNWEYWLZABJF3S5/action/author_attestation","sign_citation":"https://pith.science/pith/YLEWLVDBJ7OXNWEYWLZABJF3S5/action/citation_signature","submit_replication":"https://pith.science/pith/YLEWLVDBJ7OXNWEYWLZABJF3S5/action/replication_record"}},"created_at":"2026-07-05T11:51:19.743716+00:00","updated_at":"2026-07-05T11:51:19.743716+00:00"}