{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:LI5H22A4WHASQX7IOPEF5SF54O","short_pith_number":"pith:LI5H22A4","schema_version":"1.0","canonical_sha256":"5a3a7d681cb1c1285fe873c85ec8bde3b7182e3dccb95f71115c251822527fa6","source":{"kind":"arxiv","id":"2305.18583","version":1},"attestation_state":"computed","paper":{"title":"Controllable Text-to-Image Generation with GPT-4","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Neel Joshi, Tianjun Zhang, Vibhav Vineet, Xin Wang, Yi Zhang","submitted_at":"2023-05-29T19:56:47Z","abstract_excerpt":"Current text-to-image generation models often struggle to follow textual instructions, especially the ones requiring spatial reasoning. On the other hand, Large Language Models (LLMs), such as GPT-4, have shown remarkable precision in generating code snippets for sketching out text inputs graphically, e.g., via TikZ. In this work, we introduce Control-GPT to guide the diffusion-based text-to-image pipelines with programmatic sketches generated by GPT-4, enhancing their abilities for instruction following. Control-GPT works by querying GPT-4 to write TikZ code, and the generated sketches are us"},"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":"2305.18583","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2023-05-29T19:56:47Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"ef1bc80064d301648790ded84efd305ebe4a1aaddc50c88d3b6104f524f65484","abstract_canon_sha256":"7d8d88f2d958c1adf0e4a9528705faa84cccb455fe897826432e1280679e59d0"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:15:02.509454Z","signature_b64":"oHnyQmplf5V1B1J1T4wQw9wlKKr95ZWKFOF9u7ZSymH6k9zDfQlM9QtB6CUHeqmR8s4A7CFKryrxVsN14UYvDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5a3a7d681cb1c1285fe873c85ec8bde3b7182e3dccb95f71115c251822527fa6","last_reissued_at":"2026-07-05T06:15:02.508921Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:15:02.508921Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Controllable Text-to-Image Generation with GPT-4","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Neel Joshi, Tianjun Zhang, Vibhav Vineet, Xin Wang, Yi Zhang","submitted_at":"2023-05-29T19:56:47Z","abstract_excerpt":"Current text-to-image generation models often struggle to follow textual instructions, especially the ones requiring spatial reasoning. On the other hand, Large Language Models (LLMs), such as GPT-4, have shown remarkable precision in generating code snippets for sketching out text inputs graphically, e.g., via TikZ. In this work, we introduce Control-GPT to guide the diffusion-based text-to-image pipelines with programmatic sketches generated by GPT-4, enhancing their abilities for instruction following. Control-GPT works by querying GPT-4 to write TikZ code, and the generated sketches are us"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2305.18583","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/2305.18583/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":"2305.18583","created_at":"2026-07-05T06:15:02.508991+00:00"},{"alias_kind":"arxiv_version","alias_value":"2305.18583v1","created_at":"2026-07-05T06:15:02.508991+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2305.18583","created_at":"2026-07-05T06:15:02.508991+00:00"},{"alias_kind":"pith_short_12","alias_value":"LI5H22A4WHAS","created_at":"2026-07-05T06:15:02.508991+00:00"},{"alias_kind":"pith_short_16","alias_value":"LI5H22A4WHASQX7I","created_at":"2026-07-05T06:15:02.508991+00:00"},{"alias_kind":"pith_short_8","alias_value":"LI5H22A4","created_at":"2026-07-05T06:15:02.508991+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.26740","citing_title":"LiveEdit: Towards Real-Time Diffusion-Based Streaming Video Editing","ref_index":72,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/LI5H22A4WHASQX7IOPEF5SF54O","json":"https://pith.science/pith/LI5H22A4WHASQX7IOPEF5SF54O.json","graph_json":"https://pith.science/api/pith-number/LI5H22A4WHASQX7IOPEF5SF54O/graph.json","events_json":"https://pith.science/api/pith-number/LI5H22A4WHASQX7IOPEF5SF54O/events.json","paper":"https://pith.science/paper/LI5H22A4"},"agent_actions":{"view_html":"https://pith.science/pith/LI5H22A4WHASQX7IOPEF5SF54O","download_json":"https://pith.science/pith/LI5H22A4WHASQX7IOPEF5SF54O.json","view_paper":"https://pith.science/paper/LI5H22A4","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2305.18583&json=true","fetch_graph":"https://pith.science/api/pith-number/LI5H22A4WHASQX7IOPEF5SF54O/graph.json","fetch_events":"https://pith.science/api/pith-number/LI5H22A4WHASQX7IOPEF5SF54O/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/LI5H22A4WHASQX7IOPEF5SF54O/action/timestamp_anchor","attest_storage":"https://pith.science/pith/LI5H22A4WHASQX7IOPEF5SF54O/action/storage_attestation","attest_author":"https://pith.science/pith/LI5H22A4WHASQX7IOPEF5SF54O/action/author_attestation","sign_citation":"https://pith.science/pith/LI5H22A4WHASQX7IOPEF5SF54O/action/citation_signature","submit_replication":"https://pith.science/pith/LI5H22A4WHASQX7IOPEF5SF54O/action/replication_record"}},"created_at":"2026-07-05T06:15:02.508991+00:00","updated_at":"2026-07-05T06:15:02.508991+00:00"}