{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:XP6NOD2LBO77CO6TN47DWS5D63","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":"30dc762392598d65ed1c209eb824a4dce66ccd563822bb501edf5145cf839d7f","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-01-16T08:08:01Z","title_canon_sha256":"7ecf981aac22842ef665e3b56c0d7172dbc0ccd05e1826a59560c4e1a59c1541"},"schema_version":"1.0","source":{"id":"2401.08190","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2401.08190","created_at":"2026-07-05T07:48:02Z"},{"alias_kind":"arxiv_version","alias_value":"2401.08190v3","created_at":"2026-07-05T07:48:02Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2401.08190","created_at":"2026-07-05T07:48:02Z"},{"alias_kind":"pith_short_12","alias_value":"XP6NOD2LBO77","created_at":"2026-07-05T07:48:02Z"},{"alias_kind":"pith_short_16","alias_value":"XP6NOD2LBO77CO6T","created_at":"2026-07-05T07:48:02Z"},{"alias_kind":"pith_short_8","alias_value":"XP6NOD2L","created_at":"2026-07-05T07:48:02Z"}],"graph_snapshots":[{"event_id":"sha256:3961be57e776d57b801fec30e7971f7dd0a32c81aa15c78026314b0d9cc3a155","target":"graph","created_at":"2026-07-05T07:48:02Z","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/2401.08190/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large language models (LLMs) have seen considerable advancements in natural language understanding tasks, yet there remains a gap to bridge before attaining true artificial general intelligence, especially concerning shortcomings in mathematical reasoning capabilities. We postulate that the inherent nature of LLM training, which focuses on predicting probabilities of next token, presents challenges in effectively modeling mathematical reasoning that demands exact calculations, both from data-driven and theoretical standpoints. In this paper, we address this challenge by enriching the data land","authors_text":"Chengxi Li, Jing Wu, Kai Fan, Minpeng Liao, Wei Luo","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-01-16T08:08:01Z","title":"MARIO: MAth Reasoning with code Interpreter Output -- A Reproducible Pipeline"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2401.08190","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":"sha256:3816040a41d9826dc92f1634324ca98bb706e12f14db5450c53f59106a936f11","target":"record","created_at":"2026-07-05T07:48:02Z","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":"30dc762392598d65ed1c209eb824a4dce66ccd563822bb501edf5145cf839d7f","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-01-16T08:08:01Z","title_canon_sha256":"7ecf981aac22842ef665e3b56c0d7172dbc0ccd05e1826a59560c4e1a59c1541"},"schema_version":"1.0","source":{"id":"2401.08190","kind":"arxiv","version":3}},"canonical_sha256":"bbfcd70f4b0bbff13bd36f3e3b4ba3f6f135b9d6492d5dda7545bb04887e75a5","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"bbfcd70f4b0bbff13bd36f3e3b4ba3f6f135b9d6492d5dda7545bb04887e75a5","first_computed_at":"2026-07-05T07:48:02.719590Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T07:48:02.719590Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"OsI04AIbUmwnrU0rBDghQUT2TuvwjPl/70chFEXu/NV+zAjswBmYpSfDq26hCLhM76Xr5ZtDJQsmzU4hh5aUCQ==","signature_status":"signed_v1","signed_at":"2026-07-05T07:48:02.720125Z","signed_message":"canonical_sha256_bytes"},"source_id":"2401.08190","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:3816040a41d9826dc92f1634324ca98bb706e12f14db5450c53f59106a936f11","sha256:3961be57e776d57b801fec30e7971f7dd0a32c81aa15c78026314b0d9cc3a155"],"state_sha256":"f31c74bff8ff2cfc18cf980d3bec8f860e97819233dd1d13d48eb60945a88fd3"}