{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:727RCSC4WITGF5RZQALEJ2CLJO","short_pith_number":"pith:727RCSC4","schema_version":"1.0","canonical_sha256":"febf11485cb22662f639801644e84b4bb2d13373968eb8d71872a7064768093a","source":{"kind":"arxiv","id":"2411.17066","version":1},"attestation_state":"computed","paper":{"title":"Relations, Negations, and Numbers: Looking for Logic in Generative Text-to-Image Models","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.CL","cs.SC"],"primary_cat":"cs.CV","authors_text":"Colin Conwell, Rupert Tawiah-Quashie, Tomer Ullman","submitted_at":"2024-11-26T03:06:52Z","abstract_excerpt":"Despite remarkable progress in multi-modal AI research, there is a salient domain in which modern AI continues to lag considerably behind even human children: the reliable deployment of logical operators. Here, we examine three forms of logical operators: relations, negations, and discrete numbers. We asked human respondents (N=178 in total) to evaluate images generated by a state-of-the-art image-generating AI (DALL-E 3) prompted with these `logical probes', and find that none reliably produce human agreement scores greater than 50\\%. The negation probes and numbers (beyond 3) fail most frequ"},"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":"2411.17066","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CV","submitted_at":"2024-11-26T03:06:52Z","cross_cats_sorted":["cs.CL","cs.SC"],"title_canon_sha256":"7fa8b8acfd5e6adc9c8349e78588e850ae8fddebfdc28c69beadb1421a04dd94","abstract_canon_sha256":"ed55ca1ddd01c6ed43b0721395e5563ccd9844946f4529d7461eb08146e065a2"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:40:26.941113Z","signature_b64":"SV11BV9r1GNIHIew1wQpK5O4PWaR8ADs2GvL4ZWE93qPqeJcwJt3hDi/HaMunsw40YdPItzwLEoEKrK+O9A1Dg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"febf11485cb22662f639801644e84b4bb2d13373968eb8d71872a7064768093a","last_reissued_at":"2026-07-05T09:40:26.940358Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:40:26.940358Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Relations, Negations, and Numbers: Looking for Logic in Generative Text-to-Image Models","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.CL","cs.SC"],"primary_cat":"cs.CV","authors_text":"Colin Conwell, Rupert Tawiah-Quashie, Tomer Ullman","submitted_at":"2024-11-26T03:06:52Z","abstract_excerpt":"Despite remarkable progress in multi-modal AI research, there is a salient domain in which modern AI continues to lag considerably behind even human children: the reliable deployment of logical operators. Here, we examine three forms of logical operators: relations, negations, and discrete numbers. We asked human respondents (N=178 in total) to evaluate images generated by a state-of-the-art image-generating AI (DALL-E 3) prompted with these `logical probes', and find that none reliably produce human agreement scores greater than 50\\%. The negation probes and numbers (beyond 3) fail most frequ"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2411.17066","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/2411.17066/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":"2411.17066","created_at":"2026-07-05T09:40:26.940639+00:00"},{"alias_kind":"arxiv_version","alias_value":"2411.17066v1","created_at":"2026-07-05T09:40:26.940639+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2411.17066","created_at":"2026-07-05T09:40:26.940639+00:00"},{"alias_kind":"pith_short_12","alias_value":"727RCSC4WITG","created_at":"2026-07-05T09:40:26.940639+00:00"},{"alias_kind":"pith_short_16","alias_value":"727RCSC4WITGF5RZ","created_at":"2026-07-05T09:40:26.940639+00:00"},{"alias_kind":"pith_short_8","alias_value":"727RCSC4","created_at":"2026-07-05T09:40:26.940639+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.18943","citing_title":"Physics-IQ Verified","ref_index":34,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/727RCSC4WITGF5RZQALEJ2CLJO","json":"https://pith.science/pith/727RCSC4WITGF5RZQALEJ2CLJO.json","graph_json":"https://pith.science/api/pith-number/727RCSC4WITGF5RZQALEJ2CLJO/graph.json","events_json":"https://pith.science/api/pith-number/727RCSC4WITGF5RZQALEJ2CLJO/events.json","paper":"https://pith.science/paper/727RCSC4"},"agent_actions":{"view_html":"https://pith.science/pith/727RCSC4WITGF5RZQALEJ2CLJO","download_json":"https://pith.science/pith/727RCSC4WITGF5RZQALEJ2CLJO.json","view_paper":"https://pith.science/paper/727RCSC4","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2411.17066&json=true","fetch_graph":"https://pith.science/api/pith-number/727RCSC4WITGF5RZQALEJ2CLJO/graph.json","fetch_events":"https://pith.science/api/pith-number/727RCSC4WITGF5RZQALEJ2CLJO/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/727RCSC4WITGF5RZQALEJ2CLJO/action/timestamp_anchor","attest_storage":"https://pith.science/pith/727RCSC4WITGF5RZQALEJ2CLJO/action/storage_attestation","attest_author":"https://pith.science/pith/727RCSC4WITGF5RZQALEJ2CLJO/action/author_attestation","sign_citation":"https://pith.science/pith/727RCSC4WITGF5RZQALEJ2CLJO/action/citation_signature","submit_replication":"https://pith.science/pith/727RCSC4WITGF5RZQALEJ2CLJO/action/replication_record"}},"created_at":"2026-07-05T09:40:26.940639+00:00","updated_at":"2026-07-05T09:40:26.940639+00:00"}