{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:BMMMNBT3ULP3VBY35HVI3AIW3Z","short_pith_number":"pith:BMMMNBT3","schema_version":"1.0","canonical_sha256":"0b18c6867ba2dfba871be9ea8d8116de472aae08e2657671787359c82a3155e5","source":{"kind":"arxiv","id":"2507.05300","version":1},"attestation_state":"computed","paper":{"title":"Structured Captions Improve Prompt Adherence in Text-to-Image Models (Re-LAION-Caption 19M)","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CL","cs.LG"],"primary_cat":"cs.CV","authors_text":"Andrei Cristian Popescu, Carlos Garcia Jurado Suarez, Haitz S\\'aez de Oc\\'ariz Borde, Nicholas Merchant","submitted_at":"2025-07-07T01:18:40Z","abstract_excerpt":"We argue that generative text-to-image models often struggle with prompt adherence due to the noisy and unstructured nature of large-scale datasets like LAION-5B. This forces users to rely heavily on prompt engineering to elicit desirable outputs. In this work, we propose that enforcing a consistent caption structure during training can significantly improve model controllability and alignment. We introduce Re-LAION-Caption 19M, a high-quality subset of Re-LAION-5B, comprising 19 million 1024x1024 images with captions generated by a Mistral 7B Instruct-based LLaVA-Next model. Each caption foll"},"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":"2507.05300","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-07-07T01:18:40Z","cross_cats_sorted":["cs.AI","cs.CL","cs.LG"],"title_canon_sha256":"3d5d84ad7ebf855c8a9a125564657e23e9ba59348868a289c168c93eff33c6e6","abstract_canon_sha256":"2ac56503c832276ab2913a465932f4b5905f941b3e9b8168edb2fc7b84a78613"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:33:27.861673Z","signature_b64":"schB9DAYA8wQWSZ87ASliERRC3fOZ74We14mCqnrdc+8UnlWR9oO6YLfaP4tZwvq4iKpRHZ3B4NLlHLr79YdAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0b18c6867ba2dfba871be9ea8d8116de472aae08e2657671787359c82a3155e5","last_reissued_at":"2026-07-05T11:33:27.861254Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:33:27.861254Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Structured Captions Improve Prompt Adherence in Text-to-Image Models (Re-LAION-Caption 19M)","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CL","cs.LG"],"primary_cat":"cs.CV","authors_text":"Andrei Cristian Popescu, Carlos Garcia Jurado Suarez, Haitz S\\'aez de Oc\\'ariz Borde, Nicholas Merchant","submitted_at":"2025-07-07T01:18:40Z","abstract_excerpt":"We argue that generative text-to-image models often struggle with prompt adherence due to the noisy and unstructured nature of large-scale datasets like LAION-5B. This forces users to rely heavily on prompt engineering to elicit desirable outputs. In this work, we propose that enforcing a consistent caption structure during training can significantly improve model controllability and alignment. We introduce Re-LAION-Caption 19M, a high-quality subset of Re-LAION-5B, comprising 19 million 1024x1024 images with captions generated by a Mistral 7B Instruct-based LLaVA-Next model. Each caption foll"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2507.05300","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/2507.05300/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":"2507.05300","created_at":"2026-07-05T11:33:27.861308+00:00"},{"alias_kind":"arxiv_version","alias_value":"2507.05300v1","created_at":"2026-07-05T11:33:27.861308+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2507.05300","created_at":"2026-07-05T11:33:27.861308+00:00"},{"alias_kind":"pith_short_12","alias_value":"BMMMNBT3ULP3","created_at":"2026-07-05T11:33:27.861308+00:00"},{"alias_kind":"pith_short_16","alias_value":"BMMMNBT3ULP3VBY3","created_at":"2026-07-05T11:33:27.861308+00:00"},{"alias_kind":"pith_short_8","alias_value":"BMMMNBT3","created_at":"2026-07-05T11:33:27.861308+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2604.12357","citing_title":"ReflectCAP: Detailed Image Captioning with Reflective Memory","ref_index":24,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/BMMMNBT3ULP3VBY35HVI3AIW3Z","json":"https://pith.science/pith/BMMMNBT3ULP3VBY35HVI3AIW3Z.json","graph_json":"https://pith.science/api/pith-number/BMMMNBT3ULP3VBY35HVI3AIW3Z/graph.json","events_json":"https://pith.science/api/pith-number/BMMMNBT3ULP3VBY35HVI3AIW3Z/events.json","paper":"https://pith.science/paper/BMMMNBT3"},"agent_actions":{"view_html":"https://pith.science/pith/BMMMNBT3ULP3VBY35HVI3AIW3Z","download_json":"https://pith.science/pith/BMMMNBT3ULP3VBY35HVI3AIW3Z.json","view_paper":"https://pith.science/paper/BMMMNBT3","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2507.05300&json=true","fetch_graph":"https://pith.science/api/pith-number/BMMMNBT3ULP3VBY35HVI3AIW3Z/graph.json","fetch_events":"https://pith.science/api/pith-number/BMMMNBT3ULP3VBY35HVI3AIW3Z/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/BMMMNBT3ULP3VBY35HVI3AIW3Z/action/timestamp_anchor","attest_storage":"https://pith.science/pith/BMMMNBT3ULP3VBY35HVI3AIW3Z/action/storage_attestation","attest_author":"https://pith.science/pith/BMMMNBT3ULP3VBY35HVI3AIW3Z/action/author_attestation","sign_citation":"https://pith.science/pith/BMMMNBT3ULP3VBY35HVI3AIW3Z/action/citation_signature","submit_replication":"https://pith.science/pith/BMMMNBT3ULP3VBY35HVI3AIW3Z/action/replication_record"}},"created_at":"2026-07-05T11:33:27.861308+00:00","updated_at":"2026-07-05T11:33:27.861308+00:00"}