{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:FUZEXRTACI7KAIWFFTKTAIH62Z","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":"cd0a2c10f7780c4ded178e1f218c20b085452f754937c16e75c032cef8bee9b6","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-05-28T08:30:15Z","title_canon_sha256":"ec8ec53866eeb468e224a0743fa9dde1120d18046266f0cb4ecdda1e98e74b46"},"schema_version":"1.0","source":{"id":"2605.29586","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.29586","created_at":"2026-05-29T01:05:49Z"},{"alias_kind":"arxiv_version","alias_value":"2605.29586v1","created_at":"2026-05-29T01:05:49Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.29586","created_at":"2026-05-29T01:05:49Z"},{"alias_kind":"pith_short_12","alias_value":"FUZEXRTACI7K","created_at":"2026-05-29T01:05:49Z"},{"alias_kind":"pith_short_16","alias_value":"FUZEXRTACI7KAIWF","created_at":"2026-05-29T01:05:49Z"},{"alias_kind":"pith_short_8","alias_value":"FUZEXRTA","created_at":"2026-05-29T01:05:49Z"}],"graph_snapshots":[{"event_id":"sha256:a8992b4151f45932d19667cf96e1ac2b62e6db9584661efb6ea7ad3bbd413f18","target":"graph","created_at":"2026-05-29T01:05:49Z","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/2605.29586/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"We introduce FinVerBench, a benchmark and validity study for financial statement verification: determining whether a set of corporate financial statements is numerically consistent from the information shown to the model. FinVerBench is built from SEC 10-K XBRL filings for 43 S&P 500 companies and defines a four-category error taxonomy covering arithmetic, cross-statement linkage, year-over-year, and magnitude perturbations. We attempt fifteen contemporary LLM evaluations and report fourteen complete runs; a Gemini 2.5 Pro run is excluded from the main comparison because 40/108 gateway calls f","authors_text":"Silu Panda","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-05-28T08:30:15Z","title":"FinVerBench: Benchmark Validity and Calibration in Large Language Model Financial Statement Verification"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.29586","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":"sha256:61f693fd698364f7dc2db44e4c4cea1c1c852104f5edac8b4fe16db787357d72","target":"record","created_at":"2026-05-29T01:05:49Z","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":"cd0a2c10f7780c4ded178e1f218c20b085452f754937c16e75c032cef8bee9b6","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-05-28T08:30:15Z","title_canon_sha256":"ec8ec53866eeb468e224a0743fa9dde1120d18046266f0cb4ecdda1e98e74b46"},"schema_version":"1.0","source":{"id":"2605.29586","kind":"arxiv","version":1}},"canonical_sha256":"2d324bc660123ea022c52cd53020fed65a443b36d47118ce94e24af117f2495c","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"2d324bc660123ea022c52cd53020fed65a443b36d47118ce94e24af117f2495c","first_computed_at":"2026-05-29T01:05:49.140689Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-29T01:05:49.140689Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"ZUYWELsautYFLnCjTTlYlH3F2WltkYvJfH+gM+UIXNA9XR96u/g0nUbx86PqUxlfPH3dfi18eYeHNh2PaYVpAg==","signature_status":"signed_v1","signed_at":"2026-05-29T01:05:49.141806Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.29586","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:61f693fd698364f7dc2db44e4c4cea1c1c852104f5edac8b4fe16db787357d72","sha256:a8992b4151f45932d19667cf96e1ac2b62e6db9584661efb6ea7ad3bbd413f18"],"state_sha256":"bd64f1f48817ea2aa54b2bd885ef15cd18107519c42fc26b4a63d054d3f1816c"}