{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:JWEDMMPE46WZVG6ROTX2RF5XA2","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"1d385151d393b57ac9944aef4b86d40dfc4e96098be15b9757ff3bf704cb9f46","cross_cats_sorted":["cs.AI","cs.CL","cs.LG","cs.MM"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2024-06-19T18:00:07Z","title_canon_sha256":"987188c0d95ce261e5d58305830851b227a5aef146daec34ab5b481cc544c363"},"schema_version":"1.0","source":{"id":"2406.13743","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2406.13743","created_at":"2026-07-05T09:30:17Z"},{"alias_kind":"arxiv_version","alias_value":"2406.13743v3","created_at":"2026-07-05T09:30:17Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.13743","created_at":"2026-07-05T09:30:17Z"},{"alias_kind":"pith_short_12","alias_value":"JWEDMMPE46WZ","created_at":"2026-07-05T09:30:17Z"},{"alias_kind":"pith_short_16","alias_value":"JWEDMMPE46WZVG6R","created_at":"2026-07-05T09:30:17Z"},{"alias_kind":"pith_short_8","alias_value":"JWEDMMPE","created_at":"2026-07-05T09:30:17Z"}],"graph_snapshots":[{"event_id":"sha256:1d8d8c35f36f1219448d4d529e5741435874d6fec4a1ce7e8d0dab29441e4eb3","target":"graph","created_at":"2026-07-05T09:30:17Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2406.13743/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"While text-to-visual models now produce photo-realistic images and videos, they struggle with compositional text prompts involving attributes, relationships, and higher-order reasoning such as logic and comparison. In this work, we conduct an extensive human study on GenAI-Bench to evaluate the performance of leading image and video generation models in various aspects of compositional text-to-visual generation. We also compare automated evaluation metrics against our collected human ratings and find that VQAScore -- a metric measuring the likelihood that a VQA model views an image as accurate","authors_text":"Baiqi Li, Deepak Pathak, Deva Ramanan, Graham Neubig, Jiayao Li, Kewen Wu, Pengchuan Zhang, Tiffany Ling, Xide Xia, Yixin Fei, Zhiqiu Lin","cross_cats":["cs.AI","cs.CL","cs.LG","cs.MM"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2024-06-19T18:00:07Z","title":"GenAI-Bench: Evaluating and Improving Compositional Text-to-Visual Generation"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.13743","kind":"arxiv","version":3},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:4ae1cf2311948ca892086ee05c2d03f0a10240f9f63af82b9c1817805b06705e","target":"record","created_at":"2026-07-05T09:30:17Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"1d385151d393b57ac9944aef4b86d40dfc4e96098be15b9757ff3bf704cb9f46","cross_cats_sorted":["cs.AI","cs.CL","cs.LG","cs.MM"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2024-06-19T18:00:07Z","title_canon_sha256":"987188c0d95ce261e5d58305830851b227a5aef146daec34ab5b481cc544c363"},"schema_version":"1.0","source":{"id":"2406.13743","kind":"arxiv","version":3}},"canonical_sha256":"4d883631e4e7ad9a9bd174efa897b706b0da080abe3be91c3f9a6db367cc33f3","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"4d883631e4e7ad9a9bd174efa897b706b0da080abe3be91c3f9a6db367cc33f3","first_computed_at":"2026-07-05T09:30:17.476574Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T09:30:17.476574Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"CfTXkvsdwp48+TD3ookvqOU12hyjEQg82z10ZNz3uKj1zWD4inQB7W/fUkYTnVQpeS/14LSV5VsM8xMM1FQ2Bg==","signature_status":"signed_v1","signed_at":"2026-07-05T09:30:17.477073Z","signed_message":"canonical_sha256_bytes"},"source_id":"2406.13743","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:4ae1cf2311948ca892086ee05c2d03f0a10240f9f63af82b9c1817805b06705e","sha256:1d8d8c35f36f1219448d4d529e5741435874d6fec4a1ce7e8d0dab29441e4eb3"],"state_sha256":"0a41f6d6a18a454825318294345892388c2e74b3939ed0ef6c997c885fc9e469"}