{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:2TN5JHT4M4KXI3RG75BN2C3GB3","short_pith_number":"pith:2TN5JHT4","schema_version":"1.0","canonical_sha256":"d4dbd49e7c6715746e26ff42dd0b660eec6bece48aced007c42756da8c2806ba","source":{"kind":"arxiv","id":"2605.22885","version":1},"attestation_state":"computed","paper":{"title":"ImProver 2: Iteratively Self-Improving LMs for Neurosymbolic Proof Optimization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL","cs.LG","cs.LO"],"primary_cat":"cs.AI","authors_text":"Jeremy Avigad, Riyaz Ahuja, Sean Welleck, Tate Rowney","submitted_at":"2026-05-21T02:20:26Z","abstract_excerpt":"Formal mathematics libraries are rapidly expanding, creating a growing need to refactor verified proofs for maintainability and to improve training data quality for neural provers. However, scalable proof optimization is hindered by heterogeneous and heuristically specified objectives, scarce data, and high training and inference costs. To overcome these challenges, we introduce ImProver 2, a neurosymbolic framework for automated proof optimization in Lean 4. ImProver 2 combines a data-efficient expert-iteration pipeline with a scaffold that exposes formal structure alongside lightweight infor"},"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":"2605.22885","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-05-21T02:20:26Z","cross_cats_sorted":["cs.CL","cs.LG","cs.LO"],"title_canon_sha256":"ed5412ad9422dcdd659246d5fe216471d67d3a15ab5c5700ce097c53887a9028","abstract_canon_sha256":"5552f64973ffa2734f2f8a8e24bdd88b0bb82250162e78bf57d2a8cd4036630b"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-25T02:01:28.644967Z","signature_b64":"Qgb/WarzcTv3yfgNzxmn+jfWnwPCXT2oGtiQbSwmB/KEl3pGcm1FWVwGa70rX2MAgT6zXYFSLv5ReS9dlJzIAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d4dbd49e7c6715746e26ff42dd0b660eec6bece48aced007c42756da8c2806ba","last_reissued_at":"2026-05-25T02:01:28.644392Z","signature_status":"signed_v1","first_computed_at":"2026-05-25T02:01:28.644392Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"ImProver 2: Iteratively Self-Improving LMs for Neurosymbolic Proof Optimization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL","cs.LG","cs.LO"],"primary_cat":"cs.AI","authors_text":"Jeremy Avigad, Riyaz Ahuja, Sean Welleck, Tate Rowney","submitted_at":"2026-05-21T02:20:26Z","abstract_excerpt":"Formal mathematics libraries are rapidly expanding, creating a growing need to refactor verified proofs for maintainability and to improve training data quality for neural provers. However, scalable proof optimization is hindered by heterogeneous and heuristically specified objectives, scarce data, and high training and inference costs. To overcome these challenges, we introduce ImProver 2, a neurosymbolic framework for automated proof optimization in Lean 4. ImProver 2 combines a data-efficient expert-iteration pipeline with a scaffold that exposes formal structure alongside lightweight infor"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.22885","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/2605.22885/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":"2605.22885","created_at":"2026-05-25T02:01:28.644496+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.22885v1","created_at":"2026-05-25T02:01:28.644496+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.22885","created_at":"2026-05-25T02:01:28.644496+00:00"},{"alias_kind":"pith_short_12","alias_value":"2TN5JHT4M4KX","created_at":"2026-05-25T02:01:28.644496+00:00"},{"alias_kind":"pith_short_16","alias_value":"2TN5JHT4M4KXI3RG","created_at":"2026-05-25T02:01:28.644496+00:00"},{"alias_kind":"pith_short_8","alias_value":"2TN5JHT4","created_at":"2026-05-25T02:01:28.644496+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/2TN5JHT4M4KXI3RG75BN2C3GB3","json":"https://pith.science/pith/2TN5JHT4M4KXI3RG75BN2C3GB3.json","graph_json":"https://pith.science/api/pith-number/2TN5JHT4M4KXI3RG75BN2C3GB3/graph.json","events_json":"https://pith.science/api/pith-number/2TN5JHT4M4KXI3RG75BN2C3GB3/events.json","paper":"https://pith.science/paper/2TN5JHT4"},"agent_actions":{"view_html":"https://pith.science/pith/2TN5JHT4M4KXI3RG75BN2C3GB3","download_json":"https://pith.science/pith/2TN5JHT4M4KXI3RG75BN2C3GB3.json","view_paper":"https://pith.science/paper/2TN5JHT4","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.22885&json=true","fetch_graph":"https://pith.science/api/pith-number/2TN5JHT4M4KXI3RG75BN2C3GB3/graph.json","fetch_events":"https://pith.science/api/pith-number/2TN5JHT4M4KXI3RG75BN2C3GB3/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/2TN5JHT4M4KXI3RG75BN2C3GB3/action/timestamp_anchor","attest_storage":"https://pith.science/pith/2TN5JHT4M4KXI3RG75BN2C3GB3/action/storage_attestation","attest_author":"https://pith.science/pith/2TN5JHT4M4KXI3RG75BN2C3GB3/action/author_attestation","sign_citation":"https://pith.science/pith/2TN5JHT4M4KXI3RG75BN2C3GB3/action/citation_signature","submit_replication":"https://pith.science/pith/2TN5JHT4M4KXI3RG75BN2C3GB3/action/replication_record"}},"created_at":"2026-05-25T02:01:28.644496+00:00","updated_at":"2026-05-25T02:01:28.644496+00:00"}