{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:OSW2PBQNXX6W6QJ4K6YR45JLWB","short_pith_number":"pith:OSW2PBQN","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"},"canonical_sha256":"74ada7860dbdfd6f413c57b11e752bb059908c2e76c7ae8887562891cc9fd33d","source":{"kind":"arxiv","id":"2505.20047","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2505.20047","created_at":"2026-07-05T11:09:51Z"},{"alias_kind":"arxiv_version","alias_value":"2505.20047v1","created_at":"2026-07-05T11:09:51Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.20047","created_at":"2026-07-05T11:09:51Z"},{"alias_kind":"pith_short_12","alias_value":"OSW2PBQNXX6W","created_at":"2026-07-05T11:09:51Z"},{"alias_kind":"pith_short_16","alias_value":"OSW2PBQNXX6W6QJ4","created_at":"2026-07-05T11:09:51Z"},{"alias_kind":"pith_short_8","alias_value":"OSW2PBQN","created_at":"2026-07-05T11:09:51Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:OSW2PBQNXX6W6QJ4K6YR45JLWB","target":"record","payload":{"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"},"canonical_sha256":"74ada7860dbdfd6f413c57b11e752bb059908c2e76c7ae8887562891cc9fd33d","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"},"source_kind":"arxiv","source_id":"2505.20047","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T11:09:51Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"fRUXVJEBg6tzRP62bjkhnGr4FwyvAvw0a6s+FRYw574IlJ5iC349kLX3Ax+gE37pNekynLpFf6p9prEF3ouoAQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T01:10:29.276504Z"},"content_sha256":"77952ac793f88530c1d09df1f38a226c66ae9012f3f9bd8b6b3f2b8f0ba7d385","schema_version":"1.0","event_id":"sha256:77952ac793f88530c1d09df1f38a226c66ae9012f3f9bd8b6b3f2b8f0ba7d385"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:OSW2PBQNXX6W6QJ4K6YR45JLWB","target":"graph","payload":{"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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T11:09:51Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"ynBYlyFKowyOPR/ZDNHbPwvo1mDc+9K0wno2Jtnq9yiOVsUieN2YJb1bZtyXwd9NO2pkyO1R1SZ6KWBBZi50Dw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T01:10:29.277020Z"},"content_sha256":"1f739b4c9fc1d05b81fc6cd5abc69316eb507f9926d387c85f4755f4ff248c6b","schema_version":"1.0","event_id":"sha256:1f739b4c9fc1d05b81fc6cd5abc69316eb507f9926d387c85f4755f4ff248c6b"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/OSW2PBQNXX6W6QJ4K6YR45JLWB/bundle.json","state_url":"https://pith.science/pith/OSW2PBQNXX6W6QJ4K6YR45JLWB/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/OSW2PBQNXX6W6QJ4K6YR45JLWB/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-08-09T01:10:29Z","links":{"resolver":"https://pith.science/pith/OSW2PBQNXX6W6QJ4K6YR45JLWB","bundle":"https://pith.science/pith/OSW2PBQNXX6W6QJ4K6YR45JLWB/bundle.json","state":"https://pith.science/pith/OSW2PBQNXX6W6QJ4K6YR45JLWB/state.json","well_known_bundle":"https://pith.science/.well-known/pith/OSW2PBQNXX6W6QJ4K6YR45JLWB/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:OSW2PBQNXX6W6QJ4K6YR45JLWB","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":"669612ecc5bb0f506719960f4ea92a92badbf297fb377fd17f5c29a2c53bbbe7","cross_cats_sorted":["cs.AI","cs.LO","cs.SE"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-26T14:34:04Z","title_canon_sha256":"6f4e49681fd7c4743f22355a77c0f3de4dbc60fbc754f47a36ccc02499c9eb89"},"schema_version":"1.0","source":{"id":"2505.20047","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2505.20047","created_at":"2026-07-05T11:09:51Z"},{"alias_kind":"arxiv_version","alias_value":"2505.20047v1","created_at":"2026-07-05T11:09:51Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.20047","created_at":"2026-07-05T11:09:51Z"},{"alias_kind":"pith_short_12","alias_value":"OSW2PBQNXX6W","created_at":"2026-07-05T11:09:51Z"},{"alias_kind":"pith_short_16","alias_value":"OSW2PBQNXX6W6QJ4","created_at":"2026-07-05T11:09:51Z"},{"alias_kind":"pith_short_8","alias_value":"OSW2PBQN","created_at":"2026-07-05T11:09:51Z"}],"graph_snapshots":[{"event_id":"sha256:1f739b4c9fc1d05b81fc6cd5abc69316eb507f9926d387c85f4755f4ff248c6b","target":"graph","created_at":"2026-07-05T11:09:51Z","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/2505.20047/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"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","authors_text":"Amit Sharma, Biyao Zhang, Debargha Ganguly, Shivkumar Kalyanaraman, Sreehari Sankar, Srinivasan Iyengar, Vikash Singh, Vipin Chaudhary, Xiaotian Han, Xuecen Zhang","cross_cats":["cs.AI","cs.LO","cs.SE"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-26T14:34:04Z","title":"Grammars of Formal Uncertainty: When to Trust LLMs in Automated Reasoning Tasks"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.20047","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":"sha256:77952ac793f88530c1d09df1f38a226c66ae9012f3f9bd8b6b3f2b8f0ba7d385","target":"record","created_at":"2026-07-05T11:09:51Z","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":"669612ecc5bb0f506719960f4ea92a92badbf297fb377fd17f5c29a2c53bbbe7","cross_cats_sorted":["cs.AI","cs.LO","cs.SE"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-26T14:34:04Z","title_canon_sha256":"6f4e49681fd7c4743f22355a77c0f3de4dbc60fbc754f47a36ccc02499c9eb89"},"schema_version":"1.0","source":{"id":"2505.20047","kind":"arxiv","version":1}},"canonical_sha256":"74ada7860dbdfd6f413c57b11e752bb059908c2e76c7ae8887562891cc9fd33d","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"74ada7860dbdfd6f413c57b11e752bb059908c2e76c7ae8887562891cc9fd33d","first_computed_at":"2026-07-05T11:09:51.779479Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:09:51.779479Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"H31KhERuWCAmqKYWHT3QeYTo7TE7iX5+7i+7T+pKV+q74vqGOEF5aOM2xJtkr/HDrie6oXowUcpHkkAiGHfICw==","signature_status":"signed_v1","signed_at":"2026-07-05T11:09:51.779987Z","signed_message":"canonical_sha256_bytes"},"source_id":"2505.20047","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:77952ac793f88530c1d09df1f38a226c66ae9012f3f9bd8b6b3f2b8f0ba7d385","sha256:1f739b4c9fc1d05b81fc6cd5abc69316eb507f9926d387c85f4755f4ff248c6b"],"state_sha256":"52f71c822d7fb966fe11a30d3587e3c5a5d068260dcda81ccc43f45f81e40878"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"PvehNsurs1KYYG6/qSo5vhOXMIHjWJs1QwKKnO4CpUpc6eKF+hc4CM9gE4JL4mr9pyYnrgBj+j0IR/MuzCE7AQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-09T01:10:29.283012Z","bundle_sha256":"eff63d5154dbead89f45991829501c4ab5bdef028446a48894d942ea60c262b9"}}