{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:DLUMBWE4BSR6OBV4SBIC2M3TLT","short_pith_number":"pith:DLUMBWE4","schema_version":"1.0","canonical_sha256":"1ae8c0d89c0ca3e706bc90502d33735ccdb0d6c5109256813b73b4538275dc05","source":{"kind":"arxiv","id":"2509.15357","version":2},"attestation_state":"computed","paper":{"title":"MaskAttn-SDXL: Controllable Region-Level Text-To-Image Generation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CV","authors_text":"Anzhe Cheng, Jiahao Chen, Paul Bogdan, Yu Chang","submitted_at":"2025-09-18T18:57:47Z","abstract_excerpt":"Diffusion models have achieved strong results in text-to-image generation, but important limitations remain as prompts become more structured and multi-object. On the architecture side, U-Net backbones are efficient and stable, yet their locality makes global coordination harder, while Transformer-based diffusion models improve global interactions but at substantially higher compute and memory cost. In parallel, compositional reliability remains weak: models often mix attributes across objects, violate spatial relations, or omit requested entities, and these errors are not reliably reflected b"},"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":"2509.15357","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-09-18T18:57:47Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"4f1a04a58fe4f5b47fad1c4569a4ef42a9a3679c5f6997047a1edd2e180dc880","abstract_canon_sha256":"fcdf2c2913afbcfeca48e520aeff4c5b526862b99fcee1e529d9b2cc9b1bef3f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T00:02:06.700895Z","signature_b64":"BTelSNQWt6bdegBEylYLk6ZlXdg74dD50m6+7we7PdEFahnGBI8MTOgEgnmKh6s50oOSOrAkUiVXhcotU+5VDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"1ae8c0d89c0ca3e706bc90502d33735ccdb0d6c5109256813b73b4538275dc05","last_reissued_at":"2026-05-20T00:02:06.700245Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T00:02:06.700245Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"MaskAttn-SDXL: Controllable Region-Level Text-To-Image Generation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CV","authors_text":"Anzhe Cheng, Jiahao Chen, Paul Bogdan, Yu Chang","submitted_at":"2025-09-18T18:57:47Z","abstract_excerpt":"Diffusion models have achieved strong results in text-to-image generation, but important limitations remain as prompts become more structured and multi-object. On the architecture side, U-Net backbones are efficient and stable, yet their locality makes global coordination harder, while Transformer-based diffusion models improve global interactions but at substantially higher compute and memory cost. In parallel, compositional reliability remains weak: models often mix attributes across objects, violate spatial relations, or omit requested entities, and these errors are not reliably reflected b"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2509.15357","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/2509.15357/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":"2509.15357","created_at":"2026-05-20T00:02:06.700371+00:00"},{"alias_kind":"arxiv_version","alias_value":"2509.15357v2","created_at":"2026-05-20T00:02:06.700371+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2509.15357","created_at":"2026-05-20T00:02:06.700371+00:00"},{"alias_kind":"pith_short_12","alias_value":"DLUMBWE4BSR6","created_at":"2026-05-20T00:02:06.700371+00:00"},{"alias_kind":"pith_short_16","alias_value":"DLUMBWE4BSR6OBV4","created_at":"2026-05-20T00:02:06.700371+00:00"},{"alias_kind":"pith_short_8","alias_value":"DLUMBWE4","created_at":"2026-05-20T00:02:06.700371+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":3,"sample":[{"citing_arxiv_id":"2509.15357","citing_title":"MaskAttn-SDXL: Controllable Region-Level Text-To-Image Generation","ref_index":1,"is_internal_anchor":true},{"citing_arxiv_id":"2605.10600","citing_title":"Generate \"Normal\", Edit Poisoned: Branding Injection via Hint Embedding in Image Editing","ref_index":42,"is_internal_anchor":true},{"citing_arxiv_id":"2604.09850","citing_title":"Training-Free Object-Background Compositional T2I via Dynamic Spatial Guidance and Multi-Path Pruning","ref_index":8,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/DLUMBWE4BSR6OBV4SBIC2M3TLT","json":"https://pith.science/pith/DLUMBWE4BSR6OBV4SBIC2M3TLT.json","graph_json":"https://pith.science/api/pith-number/DLUMBWE4BSR6OBV4SBIC2M3TLT/graph.json","events_json":"https://pith.science/api/pith-number/DLUMBWE4BSR6OBV4SBIC2M3TLT/events.json","paper":"https://pith.science/paper/DLUMBWE4"},"agent_actions":{"view_html":"https://pith.science/pith/DLUMBWE4BSR6OBV4SBIC2M3TLT","download_json":"https://pith.science/pith/DLUMBWE4BSR6OBV4SBIC2M3TLT.json","view_paper":"https://pith.science/paper/DLUMBWE4","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2509.15357&json=true","fetch_graph":"https://pith.science/api/pith-number/DLUMBWE4BSR6OBV4SBIC2M3TLT/graph.json","fetch_events":"https://pith.science/api/pith-number/DLUMBWE4BSR6OBV4SBIC2M3TLT/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/DLUMBWE4BSR6OBV4SBIC2M3TLT/action/timestamp_anchor","attest_storage":"https://pith.science/pith/DLUMBWE4BSR6OBV4SBIC2M3TLT/action/storage_attestation","attest_author":"https://pith.science/pith/DLUMBWE4BSR6OBV4SBIC2M3TLT/action/author_attestation","sign_citation":"https://pith.science/pith/DLUMBWE4BSR6OBV4SBIC2M3TLT/action/citation_signature","submit_replication":"https://pith.science/pith/DLUMBWE4BSR6OBV4SBIC2M3TLT/action/replication_record"}},"created_at":"2026-05-20T00:02:06.700371+00:00","updated_at":"2026-05-20T00:02:06.700371+00:00"}