{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:7VE6WXGAA523QDXHUNHRIAQMXB","short_pith_number":"pith:7VE6WXGA","schema_version":"1.0","canonical_sha256":"fd49eb5cc00775b80ee7a34f14020cb86b1b10652998eac4a7a71d55d9a598b2","source":{"kind":"arxiv","id":"2606.31002","version":1},"attestation_state":"computed","paper":{"title":"Beyond Compilation: Evaluating Faithful Natural-Language-to-Lean Statement Formalization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL","cs.LO"],"primary_cat":"cs.AI","authors_text":"Ke Zhang, Maziar Raissi, Patricio Gallardo Candela, Sudhir Murthy, Yi Xie, Zhi Wang","submitted_at":"2026-06-30T00:27:53Z","abstract_excerpt":"Theorem-proving benchmarks evaluate proof search against fixed formal statements, but natural-language-to-Lean formalization must generate the formal statement itself. In this setting, compilation is only a validity check: a Lean declaration may type-check while omitting hypotheses, changing domains, or expressing a vacuous claim. We study faithful statement formalization as both an evaluation problem and a bottleneck-attribution problem. On a 400-entry graduate-level benchmark spanning real analysis, complex analysis, topology, and algebra, our protocol combines Lean compilation, cross-model "},"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":"2606.31002","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-06-30T00:27:53Z","cross_cats_sorted":["cs.CL","cs.LO"],"title_canon_sha256":"8720d0643cd2491d7f95a129d11b82104b2d9979fc11171ba17c473d8e41f5f8","abstract_canon_sha256":"0c63e0cd51e24efa018a76cd03f4abbeb99f1cb9aabba9417907a28955d6b054"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-01T00:17:23.926254Z","signature_b64":"nAL1Ukz81jsvVf9m1UfmlOMwS1mR4bvQFkILW5rBC+FfZJMbpiZDdqABnuIhpwVGZfhonfnNSCaCl2BHpFvGDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"fd49eb5cc00775b80ee7a34f14020cb86b1b10652998eac4a7a71d55d9a598b2","last_reissued_at":"2026-07-01T00:17:23.925776Z","signature_status":"signed_v1","first_computed_at":"2026-07-01T00:17:23.925776Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Beyond Compilation: Evaluating Faithful Natural-Language-to-Lean Statement Formalization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL","cs.LO"],"primary_cat":"cs.AI","authors_text":"Ke Zhang, Maziar Raissi, Patricio Gallardo Candela, Sudhir Murthy, Yi Xie, Zhi Wang","submitted_at":"2026-06-30T00:27:53Z","abstract_excerpt":"Theorem-proving benchmarks evaluate proof search against fixed formal statements, but natural-language-to-Lean formalization must generate the formal statement itself. In this setting, compilation is only a validity check: a Lean declaration may type-check while omitting hypotheses, changing domains, or expressing a vacuous claim. We study faithful statement formalization as both an evaluation problem and a bottleneck-attribution problem. On a 400-entry graduate-level benchmark spanning real analysis, complex analysis, topology, and algebra, our protocol combines Lean compilation, cross-model "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.31002","kind":"arxiv","version":1},"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/2606.31002/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":"2606.31002","created_at":"2026-07-01T00:17:23.925861+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.31002v1","created_at":"2026-07-01T00:17:23.925861+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.31002","created_at":"2026-07-01T00:17:23.925861+00:00"},{"alias_kind":"pith_short_12","alias_value":"7VE6WXGAA523","created_at":"2026-07-01T00:17:23.925861+00:00"},{"alias_kind":"pith_short_16","alias_value":"7VE6WXGAA523QDXH","created_at":"2026-07-01T00:17:23.925861+00:00"},{"alias_kind":"pith_short_8","alias_value":"7VE6WXGA","created_at":"2026-07-01T00:17:23.925861+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/7VE6WXGAA523QDXHUNHRIAQMXB","json":"https://pith.science/pith/7VE6WXGAA523QDXHUNHRIAQMXB.json","graph_json":"https://pith.science/api/pith-number/7VE6WXGAA523QDXHUNHRIAQMXB/graph.json","events_json":"https://pith.science/api/pith-number/7VE6WXGAA523QDXHUNHRIAQMXB/events.json","paper":"https://pith.science/paper/7VE6WXGA"},"agent_actions":{"view_html":"https://pith.science/pith/7VE6WXGAA523QDXHUNHRIAQMXB","download_json":"https://pith.science/pith/7VE6WXGAA523QDXHUNHRIAQMXB.json","view_paper":"https://pith.science/paper/7VE6WXGA","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.31002&json=true","fetch_graph":"https://pith.science/api/pith-number/7VE6WXGAA523QDXHUNHRIAQMXB/graph.json","fetch_events":"https://pith.science/api/pith-number/7VE6WXGAA523QDXHUNHRIAQMXB/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/7VE6WXGAA523QDXHUNHRIAQMXB/action/timestamp_anchor","attest_storage":"https://pith.science/pith/7VE6WXGAA523QDXHUNHRIAQMXB/action/storage_attestation","attest_author":"https://pith.science/pith/7VE6WXGAA523QDXHUNHRIAQMXB/action/author_attestation","sign_citation":"https://pith.science/pith/7VE6WXGAA523QDXHUNHRIAQMXB/action/citation_signature","submit_replication":"https://pith.science/pith/7VE6WXGAA523QDXHUNHRIAQMXB/action/replication_record"}},"created_at":"2026-07-01T00:17:23.925861+00:00","updated_at":"2026-07-01T00:17:23.925861+00:00"}