{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:OSW2PBQNXX6W6QJ4K6YR45JLWB","short_pith_number":"pith:OSW2PBQN","schema_version":"1.0","canonical_sha256":"74ada7860dbdfd6f413c57b11e752bb059908c2e76c7ae8887562891cc9fd33d","source":{"kind":"arxiv","id":"2505.20047","version":1},"attestation_state":"computed","paper":{"title":"Grammars of Formal Uncertainty: When to Trust LLMs in Automated Reasoning Tasks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LO","cs.SE"],"primary_cat":"cs.CL","authors_text":"Amit Sharma, Biyao Zhang, Debargha Ganguly, Shivkumar Kalyanaraman, Sreehari Sankar, Srinivasan Iyengar, Vikash Singh, Vipin Chaudhary, Xiaotian Han, Xuecen Zhang","submitted_at":"2025-05-26T14:34:04Z","abstract_excerpt":"Large language models (LLMs) show remarkable promise for democratizing automated reasoning by generating formal specifications. However, a fundamental tension exists: LLMs are probabilistic, while formal verification demands deterministic guarantees. This paper addresses this epistemological gap by comprehensively investigating failure modes and uncertainty quantification (UQ) in LLM-generated formal artifacts. Our systematic evaluation of five frontier LLMs reveals Satisfiability Modulo Theories (SMT) based autoformalization's domain-specific impact on accuracy (from +34.8% on logical tasks t"},"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":"2505.20047","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-26T14:34:04Z","cross_cats_sorted":["cs.AI","cs.LO","cs.SE"],"title_canon_sha256":"6f4e49681fd7c4743f22355a77c0f3de4dbc60fbc754f47a36ccc02499c9eb89","abstract_canon_sha256":"669612ecc5bb0f506719960f4ea92a92badbf297fb377fd17f5c29a2c53bbbe7"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:09:51.779987Z","signature_b64":"H31KhERuWCAmqKYWHT3QeYTo7TE7iX5+7i+7T+pKV+q74vqGOEF5aOM2xJtkr/HDrie6oXowUcpHkkAiGHfICw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"74ada7860dbdfd6f413c57b11e752bb059908c2e76c7ae8887562891cc9fd33d","last_reissued_at":"2026-07-05T11:09:51.779479Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:09:51.779479Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Grammars of Formal Uncertainty: When to Trust LLMs in Automated Reasoning Tasks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LO","cs.SE"],"primary_cat":"cs.CL","authors_text":"Amit Sharma, Biyao Zhang, Debargha Ganguly, Shivkumar Kalyanaraman, Sreehari Sankar, Srinivasan Iyengar, Vikash Singh, Vipin Chaudhary, Xiaotian Han, Xuecen Zhang","submitted_at":"2025-05-26T14:34:04Z","abstract_excerpt":"Large language models (LLMs) show remarkable promise for democratizing automated reasoning by generating formal specifications. However, a fundamental tension exists: LLMs are probabilistic, while formal verification demands deterministic guarantees. This paper addresses this epistemological gap by comprehensively investigating failure modes and uncertainty quantification (UQ) in LLM-generated formal artifacts. Our systematic evaluation of five frontier LLMs reveals Satisfiability Modulo Theories (SMT) based autoformalization's domain-specific impact on accuracy (from +34.8% on logical tasks t"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.20047","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/2505.20047/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":"2505.20047","created_at":"2026-07-05T11:09:51.779540+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.20047v1","created_at":"2026-07-05T11:09:51.779540+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.20047","created_at":"2026-07-05T11:09:51.779540+00:00"},{"alias_kind":"pith_short_12","alias_value":"OSW2PBQNXX6W","created_at":"2026-07-05T11:09:51.779540+00:00"},{"alias_kind":"pith_short_16","alias_value":"OSW2PBQNXX6W6QJ4","created_at":"2026-07-05T11:09:51.779540+00:00"},{"alias_kind":"pith_short_8","alias_value":"OSW2PBQN","created_at":"2026-07-05T11:09:51.779540+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.21928","citing_title":"CausalGuard: Conformal Inference under Graph Uncertainty","ref_index":22,"is_internal_anchor":false},{"citing_arxiv_id":"2605.14063","citing_title":"Reliability-Gated Source Anchoring for Continual Test-Time Adaptation","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2605.14063","citing_title":"Reliability-Gated Source Anchoring for Continual Test-Time Adaptation","ref_index":6,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/OSW2PBQNXX6W6QJ4K6YR45JLWB","json":"https://pith.science/pith/OSW2PBQNXX6W6QJ4K6YR45JLWB.json","graph_json":"https://pith.science/api/pith-number/OSW2PBQNXX6W6QJ4K6YR45JLWB/graph.json","events_json":"https://pith.science/api/pith-number/OSW2PBQNXX6W6QJ4K6YR45JLWB/events.json","paper":"https://pith.science/paper/OSW2PBQN"},"agent_actions":{"view_html":"https://pith.science/pith/OSW2PBQNXX6W6QJ4K6YR45JLWB","download_json":"https://pith.science/pith/OSW2PBQNXX6W6QJ4K6YR45JLWB.json","view_paper":"https://pith.science/paper/OSW2PBQN","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.20047&json=true","fetch_graph":"https://pith.science/api/pith-number/OSW2PBQNXX6W6QJ4K6YR45JLWB/graph.json","fetch_events":"https://pith.science/api/pith-number/OSW2PBQNXX6W6QJ4K6YR45JLWB/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/OSW2PBQNXX6W6QJ4K6YR45JLWB/action/timestamp_anchor","attest_storage":"https://pith.science/pith/OSW2PBQNXX6W6QJ4K6YR45JLWB/action/storage_attestation","attest_author":"https://pith.science/pith/OSW2PBQNXX6W6QJ4K6YR45JLWB/action/author_attestation","sign_citation":"https://pith.science/pith/OSW2PBQNXX6W6QJ4K6YR45JLWB/action/citation_signature","submit_replication":"https://pith.science/pith/OSW2PBQNXX6W6QJ4K6YR45JLWB/action/replication_record"}},"created_at":"2026-07-05T11:09:51.779540+00:00","updated_at":"2026-07-05T11:09:51.779540+00:00"}