{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:HKUY4TLOM6ECJ5OWBFV3DMV4Y3","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"e6dcda2b38f0162594b9f7304cef35a7927d30b592547e5b0ed465e03b8dc59f","cross_cats_sorted":["cs.AI","cs.CL","cs.PL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2026-05-26T02:12:48Z","title_canon_sha256":"2141bcede7e7dceacbb8f8e54cf3c106160632932db4588b36936a362e29b339"},"schema_version":"1.0","source":{"id":"2605.26457","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.26457","created_at":"2026-05-27T01:05:19Z"},{"alias_kind":"arxiv_version","alias_value":"2605.26457v1","created_at":"2026-05-27T01:05:19Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.26457","created_at":"2026-05-27T01:05:19Z"},{"alias_kind":"pith_short_12","alias_value":"HKUY4TLOM6EC","created_at":"2026-05-27T01:05:19Z"},{"alias_kind":"pith_short_16","alias_value":"HKUY4TLOM6ECJ5OW","created_at":"2026-05-27T01:05:19Z"},{"alias_kind":"pith_short_8","alias_value":"HKUY4TLO","created_at":"2026-05-27T01:05:19Z"}],"graph_snapshots":[{"event_id":"sha256:a640d475a6e50c667955479ba65017b36e2cf3ddccf241073c8ef56077ddac61","target":"graph","created_at":"2026-05-27T01:05:19Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2605.26457/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"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","authors_text":"Anmol Agarwal, Bryan Parno, Cedric Flamant, Jannis Limperg, Kanna Shimizu, Natalie Neamtu, Pranjal Aggarwal, Sean Welleck, Seungone Kim","cross_cats":["cs.AI","cs.CL","cs.PL"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2026-05-26T02:12:48Z","title":"Verus-SpecGym: An Agentic Environment for Evaluating Specification Autoformalization"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.26457","kind":"arxiv","version":1},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:f4ccd14c972408efad94732c6aa802a631b3473aefb2774b8708abb644c29fd6","target":"record","created_at":"2026-05-27T01:05:19Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"e6dcda2b38f0162594b9f7304cef35a7927d30b592547e5b0ed465e03b8dc59f","cross_cats_sorted":["cs.AI","cs.CL","cs.PL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2026-05-26T02:12:48Z","title_canon_sha256":"2141bcede7e7dceacbb8f8e54cf3c106160632932db4588b36936a362e29b339"},"schema_version":"1.0","source":{"id":"2605.26457","kind":"arxiv","version":1}},"canonical_sha256":"3aa98e4d6e678824f5d6096bb1b2bcc6daddcfc4567e8212fb8931e1951d3267","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"3aa98e4d6e678824f5d6096bb1b2bcc6daddcfc4567e8212fb8931e1951d3267","first_computed_at":"2026-05-27T01:05:19.023883Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-27T01:05:19.023883Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"/6EnOrcOf+lNGp2c6i6PmKCOKK/lm4qwWyvHK1wlVmKRh47idzHh8umJxO4Os4yTw5+Y0MHs7TK5pJzPK1q8CQ==","signature_status":"signed_v1","signed_at":"2026-05-27T01:05:19.024542Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.26457","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:f4ccd14c972408efad94732c6aa802a631b3473aefb2774b8708abb644c29fd6","sha256:a640d475a6e50c667955479ba65017b36e2cf3ddccf241073c8ef56077ddac61"],"state_sha256":"060e6c2e68a3888a6e960cc81d95e02b0afb32966a8cdd5e6605cfefd3ba2299"}