{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:VJMZUZAVMVVJZEZFE5E5NKMAVH","merge_version":"pith-open-graph-merge-v1","event_count":3,"valid_event_count":3,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"56bdb809f84526054d2d20f01e55aa0d494027061bfc07b559cfaf0a85d7ee23","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-07-29T14:20:13Z","title_canon_sha256":"aaf92a61aa50623ba9fd2f908b85372361d919a35e4502bf7594a0da71268d05"},"schema_version":"1.0","source":{"id":"2607.26952","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2607.26952","created_at":"2026-07-30T01:23:08Z"},{"alias_kind":"arxiv_version","alias_value":"2607.26952v1","created_at":"2026-07-30T01:23:08Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.26952","created_at":"2026-07-30T01:23:08Z"},{"alias_kind":"pith_short_12","alias_value":"VJMZUZAVMVVJ","created_at":"2026-07-30T01:23:08Z"},{"alias_kind":"pith_short_16","alias_value":"VJMZUZAVMVVJZEZF","created_at":"2026-07-30T01:23:08Z"},{"alias_kind":"pith_short_8","alias_value":"VJMZUZAV","created_at":"2026-07-30T01:23:08Z"}],"graph_snapshots":[{"event_id":"sha256:46bde1ec07f2b7fa89aea617340c57f337bcf05c9681b5547292b17a36408929","target":"graph","created_at":"2026-07-30T01:23:08Z","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/2607.26952/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"We introduce CreditCardQA, the first financial literacy benchmark for numerical reasoning derived from real credit card agreements. The dataset contains 1,800 questions, including first-person variants that reflect how consumers naturally ask about fees, interest, and payments. We evaluate a range of large language and reasoning models under Chain-of-Thought (CoT) and Program-of-Thought (PoT) prompting. Overall, PoT yields consistent performance gains, particularly for models with weaker baseline reasoning, and narrows gaps between open- and closed-source systems. Through error analysis, we sh","authors_text":"Agam Shah, Arnav Hiray, Caleb Lu, Harsit Mittal, Meghaj Tarte, Sudheer Chava","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-07-29T14:20:13Z","title":"Credit Cards, Confusion, Computation, and Consequences: What Can We Uncover About Language Model Reasoning?"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.26952","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:ce50b8c7bf326fd09b16d45aece46143a66279c3130de9f9451abd0b9007c7ab","target":"record","created_at":"2026-07-30T01:23:08Z","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":"56bdb809f84526054d2d20f01e55aa0d494027061bfc07b559cfaf0a85d7ee23","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-07-29T14:20:13Z","title_canon_sha256":"aaf92a61aa50623ba9fd2f908b85372361d919a35e4502bf7594a0da71268d05"},"schema_version":"1.0","source":{"id":"2607.26952","kind":"arxiv","version":1}},"canonical_sha256":"aa599a6415656a9c93252749d6a980a9c8ef92c2f5fb78b640921cd474c99d54","receipt":{"builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"aa599a6415656a9c93252749d6a980a9c8ef92c2f5fb78b640921cd474c99d54","first_computed_at":"2026-07-30T01:23:08.884234Z","kind":"pith_receipt","last_reissued_at":"2026-07-30T01:23:08.884234Z","receipt_version":"0.3","signature_status":"unsigned_v0"},"source_id":"2607.26952","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:ce50b8c7bf326fd09b16d45aece46143a66279c3130de9f9451abd0b9007c7ab","sha256:46bde1ec07f2b7fa89aea617340c57f337bcf05c9681b5547292b17a36408929","sha256:8f8a91bf60a21f387c0740df0f0ca6eb835c9fed7f62219569ac7aea305d160c"],"state_sha256":"7b3102aae1b641a8835ebef4932ca4dd56c05c70de249f4a051c6cab0b650e53"}