{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:HKUY4TLOM6ECJ5OWBFV3DMV4Y3","short_pith_number":"pith:HKUY4TLO","schema_version":"1.0","canonical_sha256":"3aa98e4d6e678824f5d6096bb1b2bcc6daddcfc4567e8212fb8931e1951d3267","source":{"kind":"arxiv","id":"2605.26457","version":1},"attestation_state":"computed","paper":{"title":"Verus-SpecGym: An Agentic Environment for Evaluating Specification Autoformalization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CL","cs.PL"],"primary_cat":"cs.SE","authors_text":"Anmol Agarwal, Bryan Parno, Cedric Flamant, Jannis Limperg, Kanna Shimizu, Natalie Neamtu, Pranjal Aggarwal, Sean Welleck, Seungone Kim","submitted_at":"2026-05-26T02:12:48Z","abstract_excerpt":"AI coding agents are increasingly used to write real-world software, but ensuring that their outputs are correct remains a fundamental challenge. Formal verification offers a promising path: an agent generates code together with a machine-checked proof, guaranteeing that the code satisfies a formal specification. However, there is no guarantee that the formal spec itself matches the user's intent. In this work, we study specification autoformalization: whether LLM agents can translate informal programming problems into faithful formal specifications. We introduce Verus-SpecBench, a benchmark o"},"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":"2605.26457","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2026-05-26T02:12:48Z","cross_cats_sorted":["cs.AI","cs.CL","cs.PL"],"title_canon_sha256":"2141bcede7e7dceacbb8f8e54cf3c106160632932db4588b36936a362e29b339","abstract_canon_sha256":"e6dcda2b38f0162594b9f7304cef35a7927d30b592547e5b0ed465e03b8dc59f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-27T01:05:19.024542Z","signature_b64":"/6EnOrcOf+lNGp2c6i6PmKCOKK/lm4qwWyvHK1wlVmKRh47idzHh8umJxO4Os4yTw5+Y0MHs7TK5pJzPK1q8CQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3aa98e4d6e678824f5d6096bb1b2bcc6daddcfc4567e8212fb8931e1951d3267","last_reissued_at":"2026-05-27T01:05:19.023883Z","signature_status":"signed_v1","first_computed_at":"2026-05-27T01:05:19.023883Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Verus-SpecGym: An Agentic Environment for Evaluating Specification Autoformalization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CL","cs.PL"],"primary_cat":"cs.SE","authors_text":"Anmol Agarwal, Bryan Parno, Cedric Flamant, Jannis Limperg, Kanna Shimizu, Natalie Neamtu, Pranjal Aggarwal, Sean Welleck, Seungone Kim","submitted_at":"2026-05-26T02:12:48Z","abstract_excerpt":"AI coding agents are increasingly used to write real-world software, but ensuring that their outputs are correct remains a fundamental challenge. Formal verification offers a promising path: an agent generates code together with a machine-checked proof, guaranteeing that the code satisfies a formal specification. However, there is no guarantee that the formal spec itself matches the user's intent. In this work, we study specification autoformalization: whether LLM agents can translate informal programming problems into faithful formal specifications. We introduce Verus-SpecBench, a benchmark o"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.26457","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/2605.26457/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":"2605.26457","created_at":"2026-05-27T01:05:19.023986+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.26457v1","created_at":"2026-05-27T01:05:19.023986+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.26457","created_at":"2026-05-27T01:05:19.023986+00:00"},{"alias_kind":"pith_short_12","alias_value":"HKUY4TLOM6EC","created_at":"2026-05-27T01:05:19.023986+00:00"},{"alias_kind":"pith_short_16","alias_value":"HKUY4TLOM6ECJ5OW","created_at":"2026-05-27T01:05:19.023986+00:00"},{"alias_kind":"pith_short_8","alias_value":"HKUY4TLO","created_at":"2026-05-27T01:05:19.023986+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/HKUY4TLOM6ECJ5OWBFV3DMV4Y3","json":"https://pith.science/pith/HKUY4TLOM6ECJ5OWBFV3DMV4Y3.json","graph_json":"https://pith.science/api/pith-number/HKUY4TLOM6ECJ5OWBFV3DMV4Y3/graph.json","events_json":"https://pith.science/api/pith-number/HKUY4TLOM6ECJ5OWBFV3DMV4Y3/events.json","paper":"https://pith.science/paper/HKUY4TLO"},"agent_actions":{"view_html":"https://pith.science/pith/HKUY4TLOM6ECJ5OWBFV3DMV4Y3","download_json":"https://pith.science/pith/HKUY4TLOM6ECJ5OWBFV3DMV4Y3.json","view_paper":"https://pith.science/paper/HKUY4TLO","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.26457&json=true","fetch_graph":"https://pith.science/api/pith-number/HKUY4TLOM6ECJ5OWBFV3DMV4Y3/graph.json","fetch_events":"https://pith.science/api/pith-number/HKUY4TLOM6ECJ5OWBFV3DMV4Y3/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/HKUY4TLOM6ECJ5OWBFV3DMV4Y3/action/timestamp_anchor","attest_storage":"https://pith.science/pith/HKUY4TLOM6ECJ5OWBFV3DMV4Y3/action/storage_attestation","attest_author":"https://pith.science/pith/HKUY4TLOM6ECJ5OWBFV3DMV4Y3/action/author_attestation","sign_citation":"https://pith.science/pith/HKUY4TLOM6ECJ5OWBFV3DMV4Y3/action/citation_signature","submit_replication":"https://pith.science/pith/HKUY4TLOM6ECJ5OWBFV3DMV4Y3/action/replication_record"}},"created_at":"2026-05-27T01:05:19.023986+00:00","updated_at":"2026-05-27T01:05:19.023986+00:00"}