{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:P2FNYCMTM543BVY6F52Z4IVN4S","short_pith_number":"pith:P2FNYCMT","schema_version":"1.0","canonical_sha256":"7e8adc09936779b0d71e2f759e22ade4b0af713f39017259c58e9a7bc13daa4d","source":{"kind":"arxiv","id":"2605.14061","version":1},"attestation_state":"computed","paper":{"title":"MathAtlas: A Benchmark for Autoformalization in the Wild","license":"http://creativecommons.org/licenses/by/4.0/","headline":"MathAtlas extracts 52k graduate-level math statements from textbooks to benchmark autoformalization systems.","cross_cats":["cs.LG"],"primary_cat":"cs.AI","authors_text":"Davit Babayan, Hafsah Mahmood, Jeffrey Flanigan, Laurel Willey, Liam McCarty, Nilay Patel, Noah Arias, Soli Munoz, Timothy Libman, Victoria Cochran","submitted_at":"2026-05-13T19:35:46Z","abstract_excerpt":"Current autoformalization benchmarks are largely focused on olympiad or undergraduate mathematics, while graduate and research-level mathematics remains underexplored. In this paper, we introduce MathAtlas, the first large-scale autoformalization benchmark of in the wild graduate-level mathematics, containing ~52k theorems, definitions, exercises, examples, and proofs extracted from 103 graduate mathematics textbooks. MathAtlas is enriched with a mathematical dependency graph containing ~178k relations, and is the first autoformalization benchmark to include such relations, facilitating evalua"},"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":true,"formal_links_present":true},"canonical_record":{"source":{"id":"2605.14061","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-05-13T19:35:46Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"548d1dde5b807c3b1ea3382ef7b1edd6616e5abd2f965726c9b4f53db2a60df1","abstract_canon_sha256":"74d2fc61ad397429c54a53a3c8c953b1ddb337746c9c47b077e90ee7cc72012b"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:39:12.527195Z","signature_b64":"uhml0GiImbPbH66OfTfiN7OHVaoN45HjSwOfxGtlh4x89FO1yOVePwwMnFn3pLHUfDxY7WQnwOZf23/8GgiyCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7e8adc09936779b0d71e2f759e22ade4b0af713f39017259c58e9a7bc13daa4d","last_reissued_at":"2026-05-17T23:39:12.526508Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:39:12.526508Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"MathAtlas: A Benchmark for Autoformalization in the Wild","license":"http://creativecommons.org/licenses/by/4.0/","headline":"MathAtlas extracts 52k graduate-level math statements from textbooks to benchmark autoformalization systems.","cross_cats":["cs.LG"],"primary_cat":"cs.AI","authors_text":"Davit Babayan, Hafsah Mahmood, Jeffrey Flanigan, Laurel Willey, Liam McCarty, Nilay Patel, Noah Arias, Soli Munoz, Timothy Libman, Victoria Cochran","submitted_at":"2026-05-13T19:35:46Z","abstract_excerpt":"Current autoformalization benchmarks are largely focused on olympiad or undergraduate mathematics, while graduate and research-level mathematics remains underexplored. In this paper, we introduce MathAtlas, the first large-scale autoformalization benchmark of in the wild graduate-level mathematics, containing ~52k theorems, definitions, exercises, examples, and proofs extracted from 103 graduate mathematics textbooks. MathAtlas is enriched with a mathematical dependency graph containing ~178k relations, and is the first autoformalization benchmark to include such relations, facilitating evalua"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"MathAtlas is high quality but extremely challenging: strong baselines achieve at most 9.8% correctness on theorem statements and 16.7% on definitions. On MA-Hard, the best model achieves only 2.6% correctness.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the automatic extraction process from textbooks produces accurate, correctly scoped graduate-level statements and that the constructed dependency graph faithfully reflects mathematical prerequisites without introducing extraction errors.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"MathAtlas is the first large-scale benchmark for autoformalizing graduate mathematics, where even strong models reach only 9.8% correctness on theorem statements and drop to 2.6% on the hardest dependency-deep subset.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"MathAtlas extracts 52k graduate-level math statements from textbooks to benchmark autoformalization systems.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"d63b2002e535cff6824af712a6325f7b6cf6c1f9df0e5e2a20d25399a1268c83"},"source":{"id":"2605.14061","kind":"arxiv","version":1},"verdict":{"id":"2aff27b4-fc4c-44b0-85a7-d3e3954aa53b","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-15T05:05:48.174131Z","strongest_claim":"MathAtlas is high quality but extremely challenging: strong baselines achieve at most 9.8% correctness on theorem statements and 16.7% on definitions. On MA-Hard, the best model achieves only 2.6% correctness.","one_line_summary":"MathAtlas is the first large-scale benchmark for autoformalizing graduate mathematics, where even strong models reach only 9.8% correctness on theorem statements and drop to 2.6% on the hardest dependency-deep subset.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the automatic extraction process from textbooks produces accurate, correctly scoped graduate-level statements and that the constructed dependency graph faithfully reflects mathematical prerequisites without introducing extraction errors.","pith_extraction_headline":"MathAtlas extracts 52k graduate-level math statements from textbooks to benchmark autoformalization systems."},"references":{"count":54,"sample":[{"doi":"10.1145/3372885.3373827","year":null,"title":"The Lean Mathematical Library","work_id":"abd0e478-2994-4be9-8cf1-5cf6fe7ef982","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"A. Agrawal, S. Gadgil, N. Goyal, A. Narayanan, and A. Tadipatri. Towards a Mathematics Formalisation Assistant using Large Language Models. URL http://arxiv.org/abs/2211. 07524","work_id":"38d80bc0-7b4c-4475-9897-5684af27151c","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"Z. Azerbayev, B. Piotrowski, and J. Avigad. ProofNet: A Benchmark for Autoformalizing and Formally Proving Undergraduate-Level Mathematics Problems","work_id":"7cb35535-66ce-4049-b00a-d692a32c02da","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2023,"title":"Z. Azerbayev, H. Schoelkopf, K. Paster, M. D. Santos, S. McAleer, A. Q. Jiang, J. Deng, S. Biderman, and S. Welleck. Llemma: An Open Language Model For Mathematics, Oct. 2023","work_id":"d94176ed-4180-4556-b218-49f8101b619d","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2023,"title":"L. Blecher, G. Cucurull, T. Scialom, and R. Stojnic. Nougat: Neural optical understanding for academic documents, 2023","work_id":"9acde2c6-2d5d-4706-8e69-874a806f4d42","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":54,"snapshot_sha256":"6eb1b57f08b773f311c5ff1fe1893ee52c99f4cc9bc25b1c4aaa64d57dbda572","internal_anchors":3},"formal_canon":{"evidence_count":2,"snapshot_sha256":"3bafcdfb58ceaf5e3ba57a8b6a034711d287826c3eb9dee6f63f4ae76966ae7f"},"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":"2605.14061","created_at":"2026-05-17T23:39:12.526634+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.14061v1","created_at":"2026-05-17T23:39:12.526634+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.14061","created_at":"2026-05-17T23:39:12.526634+00:00"},{"alias_kind":"pith_short_12","alias_value":"P2FNYCMTM543","created_at":"2026-05-18T12:33:37.589309+00:00"},{"alias_kind":"pith_short_16","alias_value":"P2FNYCMTM543BVY6","created_at":"2026-05-18T12:33:37.589309+00:00"},{"alias_kind":"pith_short_8","alias_value":"P2FNYCMT","created_at":"2026-05-18T12:33:37.589309+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":2,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/P2FNYCMTM543BVY6F52Z4IVN4S","json":"https://pith.science/pith/P2FNYCMTM543BVY6F52Z4IVN4S.json","graph_json":"https://pith.science/api/pith-number/P2FNYCMTM543BVY6F52Z4IVN4S/graph.json","events_json":"https://pith.science/api/pith-number/P2FNYCMTM543BVY6F52Z4IVN4S/events.json","paper":"https://pith.science/paper/P2FNYCMT"},"agent_actions":{"view_html":"https://pith.science/pith/P2FNYCMTM543BVY6F52Z4IVN4S","download_json":"https://pith.science/pith/P2FNYCMTM543BVY6F52Z4IVN4S.json","view_paper":"https://pith.science/paper/P2FNYCMT","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.14061&json=true","fetch_graph":"https://pith.science/api/pith-number/P2FNYCMTM543BVY6F52Z4IVN4S/graph.json","fetch_events":"https://pith.science/api/pith-number/P2FNYCMTM543BVY6F52Z4IVN4S/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/P2FNYCMTM543BVY6F52Z4IVN4S/action/timestamp_anchor","attest_storage":"https://pith.science/pith/P2FNYCMTM543BVY6F52Z4IVN4S/action/storage_attestation","attest_author":"https://pith.science/pith/P2FNYCMTM543BVY6F52Z4IVN4S/action/author_attestation","sign_citation":"https://pith.science/pith/P2FNYCMTM543BVY6F52Z4IVN4S/action/citation_signature","submit_replication":"https://pith.science/pith/P2FNYCMTM543BVY6F52Z4IVN4S/action/replication_record"}},"created_at":"2026-05-17T23:39:12.526634+00:00","updated_at":"2026-05-17T23:39:12.526634+00:00"}