{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:HFNHRKHU6W4BDTPNHBQPDW7ENF","short_pith_number":"pith:HFNHRKHU","schema_version":"1.0","canonical_sha256":"395a78a8f4f5b811cded3860f1dbe4696b23d85c58ca255bd1167d40f63ceddd","source":{"kind":"arxiv","id":"2403.06381","version":1},"attestation_state":"computed","paper":{"title":"Enhancing Semantic Fidelity in Text-to-Image Synthesis: Attention Regulation in Diffusion Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Kenji Kawaguchi, Lim Wei Hern, Teoh Tze Tzun, Tiviatis Sim, Yang Zhang","submitted_at":"2024-03-11T02:18:27Z","abstract_excerpt":"Recent advancements in diffusion models have notably improved the perceptual quality of generated images in text-to-image synthesis tasks. However, diffusion models often struggle to produce images that accurately reflect the intended semantics of the associated text prompts. We examine cross-attention layers in diffusion models and observe a propensity for these layers to disproportionately focus on certain tokens during the generation process, thereby undermining semantic fidelity. To address the issue of dominant attention, we introduce attention regulation, a computation-efficient on-the-f"},"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":"2403.06381","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2024-03-11T02:18:27Z","cross_cats_sorted":[],"title_canon_sha256":"a51b862068bbc1cccf0f59a226ae5cffef9a8ed7c4063226a27d6eb409d9fd86","abstract_canon_sha256":"d55ce7f98d515cdab0dc7336bbf92bc676713098714f998bad442d14d6c54a9c"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:54:22.716525Z","signature_b64":"OGCTtJqjqEth2HWTsR/sLYE26ytzMTu1J6yHLkYs1+HqW/J2jwyTGfjGup/2CojnpuWAi76alWcghiKX9RjyCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"395a78a8f4f5b811cded3860f1dbe4696b23d85c58ca255bd1167d40f63ceddd","last_reissued_at":"2026-07-05T07:54:22.716031Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:54:22.716031Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Enhancing Semantic Fidelity in Text-to-Image Synthesis: Attention Regulation in Diffusion Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Kenji Kawaguchi, Lim Wei Hern, Teoh Tze Tzun, Tiviatis Sim, Yang Zhang","submitted_at":"2024-03-11T02:18:27Z","abstract_excerpt":"Recent advancements in diffusion models have notably improved the perceptual quality of generated images in text-to-image synthesis tasks. However, diffusion models often struggle to produce images that accurately reflect the intended semantics of the associated text prompts. We examine cross-attention layers in diffusion models and observe a propensity for these layers to disproportionately focus on certain tokens during the generation process, thereby undermining semantic fidelity. To address the issue of dominant attention, we introduce attention regulation, a computation-efficient on-the-f"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2403.06381","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/2403.06381/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":"2403.06381","created_at":"2026-07-05T07:54:22.716099+00:00"},{"alias_kind":"arxiv_version","alias_value":"2403.06381v1","created_at":"2026-07-05T07:54:22.716099+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2403.06381","created_at":"2026-07-05T07:54:22.716099+00:00"},{"alias_kind":"pith_short_12","alias_value":"HFNHRKHU6W4B","created_at":"2026-07-05T07:54:22.716099+00:00"},{"alias_kind":"pith_short_16","alias_value":"HFNHRKHU6W4BDTPN","created_at":"2026-07-05T07:54:22.716099+00:00"},{"alias_kind":"pith_short_8","alias_value":"HFNHRKHU","created_at":"2026-07-05T07:54:22.716099+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2509.17458","citing_title":"CARINOX: Inference-time Scaling with Category-Aware Reward-based Initial Noise Optimization and Exploration","ref_index":25,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/HFNHRKHU6W4BDTPNHBQPDW7ENF","json":"https://pith.science/pith/HFNHRKHU6W4BDTPNHBQPDW7ENF.json","graph_json":"https://pith.science/api/pith-number/HFNHRKHU6W4BDTPNHBQPDW7ENF/graph.json","events_json":"https://pith.science/api/pith-number/HFNHRKHU6W4BDTPNHBQPDW7ENF/events.json","paper":"https://pith.science/paper/HFNHRKHU"},"agent_actions":{"view_html":"https://pith.science/pith/HFNHRKHU6W4BDTPNHBQPDW7ENF","download_json":"https://pith.science/pith/HFNHRKHU6W4BDTPNHBQPDW7ENF.json","view_paper":"https://pith.science/paper/HFNHRKHU","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2403.06381&json=true","fetch_graph":"https://pith.science/api/pith-number/HFNHRKHU6W4BDTPNHBQPDW7ENF/graph.json","fetch_events":"https://pith.science/api/pith-number/HFNHRKHU6W4BDTPNHBQPDW7ENF/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/HFNHRKHU6W4BDTPNHBQPDW7ENF/action/timestamp_anchor","attest_storage":"https://pith.science/pith/HFNHRKHU6W4BDTPNHBQPDW7ENF/action/storage_attestation","attest_author":"https://pith.science/pith/HFNHRKHU6W4BDTPNHBQPDW7ENF/action/author_attestation","sign_citation":"https://pith.science/pith/HFNHRKHU6W4BDTPNHBQPDW7ENF/action/citation_signature","submit_replication":"https://pith.science/pith/HFNHRKHU6W4BDTPNHBQPDW7ENF/action/replication_record"}},"created_at":"2026-07-05T07:54:22.716099+00:00","updated_at":"2026-07-05T07:54:22.716099+00:00"}