{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:3S4XIRWMWZQC3S2ERPNGF4QKEZ","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":"a6cc1733f27bee745fcd0306fce3ef0ccb8fa2be8215bf4d5d3fecb79e5fac3b","cross_cats_sorted":["cs.AI","cs.CV"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-03-30T20:03:36Z","title_canon_sha256":"61af7160a42f1a49b28f3dc690f0eb9621f90d840ba212e728bf81431de58276"},"schema_version":"1.0","source":{"id":"2504.00043","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2504.00043","created_at":"2026-07-05T11:52:19Z"},{"alias_kind":"arxiv_version","alias_value":"2504.00043v2","created_at":"2026-07-05T11:52:19Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2504.00043","created_at":"2026-07-05T11:52:19Z"},{"alias_kind":"pith_short_12","alias_value":"3S4XIRWMWZQC","created_at":"2026-07-05T11:52:19Z"},{"alias_kind":"pith_short_16","alias_value":"3S4XIRWMWZQC3S2E","created_at":"2026-07-05T11:52:19Z"},{"alias_kind":"pith_short_8","alias_value":"3S4XIRWM","created_at":"2026-07-05T11:52:19Z"}],"graph_snapshots":[{"event_id":"sha256:7d130f7fb180e9af87e2c2c3f07e6607e0a1ecaaaf37c63a537c5e3a820b38e2","target":"graph","created_at":"2026-07-05T11:52:19Z","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/2504.00043/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Existing reasoning evaluation frameworks for Large Language Models (LLMs) and Large Vision-Language Models (LVLMs) predominantly assess either text-based reasoning or vision-language understanding capabilities, with limited dynamic interplay between textual and visual constraints. To address this limitation, we introduce CrossWordBench, a benchmark designed to evaluate the reasoning capabilities of both LLMs and LVLMs through the medium of crossword puzzles -- a task requiring multimodal adherence to semantic constraints from text-based clues and intersectional constraints from visual grid str","authors_text":"Bill Yuchen Lin, Chengsong Huang, Haohan Wang, Jiaxin Huang, Jixuan Leng, Langlin Huang, William W. Cohen","cross_cats":["cs.AI","cs.CV"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-03-30T20:03:36Z","title":"CrossWordBench: Evaluating the Reasoning Capabilities of LLMs and LVLMs with Controllable Puzzle Generation"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2504.00043","kind":"arxiv","version":2},"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:a2a7d01dc51034f6b3ace5cb533e704c9638dd8d98fb9449d31afad16cef841f","target":"record","created_at":"2026-07-05T11:52:19Z","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":"a6cc1733f27bee745fcd0306fce3ef0ccb8fa2be8215bf4d5d3fecb79e5fac3b","cross_cats_sorted":["cs.AI","cs.CV"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-03-30T20:03:36Z","title_canon_sha256":"61af7160a42f1a49b28f3dc690f0eb9621f90d840ba212e728bf81431de58276"},"schema_version":"1.0","source":{"id":"2504.00043","kind":"arxiv","version":2}},"canonical_sha256":"dcb97446ccb6602dcb448bda62f20a266dbdc3d0255bd5587bcabf0e29684427","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"dcb97446ccb6602dcb448bda62f20a266dbdc3d0255bd5587bcabf0e29684427","first_computed_at":"2026-07-05T11:52:19.306683Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:52:19.306683Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"rsQdGGrm6iOkqh5VMijsypku/TFvaLJ5YbJf5UAfDKj6WdBOJYnf7aBlCW+RIXE4IJuA+sX12reF0Qtavqe0Cw==","signature_status":"signed_v1","signed_at":"2026-07-05T11:52:19.307190Z","signed_message":"canonical_sha256_bytes"},"source_id":"2504.00043","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:a2a7d01dc51034f6b3ace5cb533e704c9638dd8d98fb9449d31afad16cef841f","sha256:7d130f7fb180e9af87e2c2c3f07e6607e0a1ecaaaf37c63a537c5e3a820b38e2"],"state_sha256":"4ef5640f1d61515ce52128c2eb8689403a87c4ff8c36a76ba5748a8a6b46dc5b"}