{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:PNSVU4Q6ZB7TDBYRH5KNM6I5R3","short_pith_number":"pith:PNSVU4Q6","schema_version":"1.0","canonical_sha256":"7b655a721ec87f3187113f54d6791d8ef263cbca3ce26f7f456db3cec1839066","source":{"kind":"arxiv","id":"2502.11574","version":2},"attestation_state":"computed","paper":{"title":"Large Language Models and Mathematical Reasoning Failures","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Birger Moell, Johan Boye","submitted_at":"2025-02-17T09:07:32Z","abstract_excerpt":"This paper investigates the mathematical reasoning capabilities of large language models (LLMs) using 50 newly constructed high-school-level word problems. Unlike prior studies that focus solely on answer correctness, we rigorously analyze both final answers and solution steps to identify reasoning failures. Evaluating eight state-of-the-art models - including Mixtral, Llama, Gemini, GPT-4o, and OpenAI's o1 variants - we find that while newer models (e.g., o3-mini, deepseek-r1) achieve higher accuracy, all models exhibit errors in spatial reasoning, strategic planning, and arithmetic, sometime"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2502.11574","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2025-02-17T09:07:32Z","cross_cats_sorted":[],"title_canon_sha256":"fdcc2bc744f5f316dbbd71e9adaf0020567f0528ba03bd729edb7834f2baa6c1","abstract_canon_sha256":"46aacc707165d16b1a9926c1ab3a10d163d1d8f6b9b3a6347f5e533c03fe021e"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:17:54.450219Z","signature_b64":"8zio13q0qxBGXzOfgRcJvLVJcPq8/ELfdNMfGhgNG2UsltG1BqUaO9ri7g52PTpZ4+q8+SsFcsACt5VNqvtAAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7b655a721ec87f3187113f54d6791d8ef263cbca3ce26f7f456db3cec1839066","last_reissued_at":"2026-07-05T10:17:54.449790Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:17:54.449790Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Large Language Models and Mathematical Reasoning Failures","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Birger Moell, Johan Boye","submitted_at":"2025-02-17T09:07:32Z","abstract_excerpt":"This paper investigates the mathematical reasoning capabilities of large language models (LLMs) using 50 newly constructed high-school-level word problems. Unlike prior studies that focus solely on answer correctness, we rigorously analyze both final answers and solution steps to identify reasoning failures. Evaluating eight state-of-the-art models - including Mixtral, Llama, Gemini, GPT-4o, and OpenAI's o1 variants - we find that while newer models (e.g., o3-mini, deepseek-r1) achieve higher accuracy, all models exhibit errors in spatial reasoning, strategic planning, and arithmetic, sometime"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.11574","kind":"arxiv","version":2},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2502.11574/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2502.11574","created_at":"2026-07-05T10:17:54.449839+00:00"},{"alias_kind":"arxiv_version","alias_value":"2502.11574v2","created_at":"2026-07-05T10:17:54.449839+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.11574","created_at":"2026-07-05T10:17:54.449839+00:00"},{"alias_kind":"pith_short_12","alias_value":"PNSVU4Q6ZB7T","created_at":"2026-07-05T10:17:54.449839+00:00"},{"alias_kind":"pith_short_16","alias_value":"PNSVU4Q6ZB7TDBYR","created_at":"2026-07-05T10:17:54.449839+00:00"},{"alias_kind":"pith_short_8","alias_value":"PNSVU4Q6","created_at":"2026-07-05T10:17:54.449839+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.02035","citing_title":"VIDA: A dataset for Visually Dependent Ambiguity in Multimodal Machine Translation","ref_index":44,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/PNSVU4Q6ZB7TDBYRH5KNM6I5R3","json":"https://pith.science/pith/PNSVU4Q6ZB7TDBYRH5KNM6I5R3.json","graph_json":"https://pith.science/api/pith-number/PNSVU4Q6ZB7TDBYRH5KNM6I5R3/graph.json","events_json":"https://pith.science/api/pith-number/PNSVU4Q6ZB7TDBYRH5KNM6I5R3/events.json","paper":"https://pith.science/paper/PNSVU4Q6"},"agent_actions":{"view_html":"https://pith.science/pith/PNSVU4Q6ZB7TDBYRH5KNM6I5R3","download_json":"https://pith.science/pith/PNSVU4Q6ZB7TDBYRH5KNM6I5R3.json","view_paper":"https://pith.science/paper/PNSVU4Q6","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2502.11574&json=true","fetch_graph":"https://pith.science/api/pith-number/PNSVU4Q6ZB7TDBYRH5KNM6I5R3/graph.json","fetch_events":"https://pith.science/api/pith-number/PNSVU4Q6ZB7TDBYRH5KNM6I5R3/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/PNSVU4Q6ZB7TDBYRH5KNM6I5R3/action/timestamp_anchor","attest_storage":"https://pith.science/pith/PNSVU4Q6ZB7TDBYRH5KNM6I5R3/action/storage_attestation","attest_author":"https://pith.science/pith/PNSVU4Q6ZB7TDBYRH5KNM6I5R3/action/author_attestation","sign_citation":"https://pith.science/pith/PNSVU4Q6ZB7TDBYRH5KNM6I5R3/action/citation_signature","submit_replication":"https://pith.science/pith/PNSVU4Q6ZB7TDBYRH5KNM6I5R3/action/replication_record"}},"created_at":"2026-07-05T10:17:54.449839+00:00","updated_at":"2026-07-05T10:17:54.449839+00:00"}