{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:3QP63GBA6B5R3TX2BIO4SB3AH6","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":"9ff1355e3e3601bcb271619e92083a4d5ec5dd8542e9f3407b40c5bc8aa1532b","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CV","submitted_at":"2026-06-04T01:11:21Z","title_canon_sha256":"04942ebe2309323f566c97e9c70a913a83edd81fdcde22b166ae483b9a592874"},"schema_version":"1.0","source":{"id":"2606.06538","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2606.06538","created_at":"2026-06-08T00:03:41Z"},{"alias_kind":"arxiv_version","alias_value":"2606.06538v1","created_at":"2026-06-08T00:03:41Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.06538","created_at":"2026-06-08T00:03:41Z"},{"alias_kind":"pith_short_12","alias_value":"3QP63GBA6B5R","created_at":"2026-06-08T00:03:41Z"},{"alias_kind":"pith_short_16","alias_value":"3QP63GBA6B5R3TX2","created_at":"2026-06-08T00:03:41Z"},{"alias_kind":"pith_short_8","alias_value":"3QP63GBA","created_at":"2026-06-08T00:03:41Z"}],"graph_snapshots":[{"event_id":"sha256:76603b0ab0a465e557f14aad1e8e2d86c350b317d8d8cfe73560e7536bfd6593","target":"graph","created_at":"2026-06-08T00:03:41Z","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/2606.06538/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"In real-world applications, models are expected to perform reliably across diverse settings. Yet, many existing multimodal benchmarks expand task types without capturing the visual diversity needed to handle open-ended visual inputs. We present WorldBench, a challenging and visually diverse reasoning benchmark to evaluate Multimodal Large Language Models (MLLMs). We build a taxonomy of thousands of visual concepts across multiple domains (e.g., living things). Guided by this taxonomy, we curate a broad collection of images from search engines and existing datasets to comprehensively represent ","authors_text":"Aleksandra Korolova, Boya Zeng, Chung Peng Lee, Gabriel Sarch, Harish Krishnakumar, Hu Xu, Shengbang Tong, Wenhao Chai, Wenhu Chen, Xingyu Fu, Yida Yin, Zhuang Liu","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CV","submitted_at":"2026-06-04T01:11:21Z","title":"WorldBench: A Challenging and Visually Diverse Multimodal Reasoning Benchmark"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.06538","kind":"arxiv","version":1},"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":"sha256:93127217bb8cda41f5518fddfed6b744ca8b8241d4eaa8f08eadab266766438f","target":"record","created_at":"2026-06-08T00:03:41Z","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":"9ff1355e3e3601bcb271619e92083a4d5ec5dd8542e9f3407b40c5bc8aa1532b","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CV","submitted_at":"2026-06-04T01:11:21Z","title_canon_sha256":"04942ebe2309323f566c97e9c70a913a83edd81fdcde22b166ae483b9a592874"},"schema_version":"1.0","source":{"id":"2606.06538","kind":"arxiv","version":1}},"canonical_sha256":"dc1fed9820f07b1dcefa0a1dc907603fb61ebca57a1f44d0d00a1e945b2543ec","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"dc1fed9820f07b1dcefa0a1dc907603fb61ebca57a1f44d0d00a1e945b2543ec","first_computed_at":"2026-06-08T00:03:41.595363Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-08T00:03:41.595363Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"kDjVk4jHtlk0jnP5jjUkC8P4sOLFmfT/0Do3o7n0Mh5K60ep1RBGWyJjMpZGb4hnOO66od2aLop92QPhhaluBA==","signature_status":"signed_v1","signed_at":"2026-06-08T00:03:41.596110Z","signed_message":"canonical_sha256_bytes"},"source_id":"2606.06538","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:93127217bb8cda41f5518fddfed6b744ca8b8241d4eaa8f08eadab266766438f","sha256:76603b0ab0a465e557f14aad1e8e2d86c350b317d8d8cfe73560e7536bfd6593"],"state_sha256":"74204fd7f3f179bc3ea234466e9b4539052f4d4710117335cfea17db4cad55ce"}