{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:ZJYDWI4EAR2UZQ5GM7PX44PR6M","short_pith_number":"pith:ZJYDWI4E","canonical_record":{"source":{"id":"2503.04779","kind":"arxiv","version":4},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.PL","submitted_at":"2025-02-22T13:27:31Z","cross_cats_sorted":["cs.AI","cs.SE"],"title_canon_sha256":"026ae5ab35cfe9c787605c1d4cb9b76ce5518277ca370105283928d16e449a81","abstract_canon_sha256":"19298fead70b53b2494bd617c6b06da35123df2c2e325daf49b08d48ed24e555"},"schema_version":"1.0"},"canonical_sha256":"ca703b238404754cc3a667df7e71f1f333c44750d97876813e7d730edba83e0b","source":{"kind":"arxiv","id":"2503.04779","version":4},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2503.04779","created_at":"2026-07-05T11:11:42Z"},{"alias_kind":"arxiv_version","alias_value":"2503.04779v4","created_at":"2026-07-05T11:11:42Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.04779","created_at":"2026-07-05T11:11:42Z"},{"alias_kind":"pith_short_12","alias_value":"ZJYDWI4EAR2U","created_at":"2026-07-05T11:11:42Z"},{"alias_kind":"pith_short_16","alias_value":"ZJYDWI4EAR2UZQ5G","created_at":"2026-07-05T11:11:42Z"},{"alias_kind":"pith_short_8","alias_value":"ZJYDWI4E","created_at":"2026-07-05T11:11:42Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:ZJYDWI4EAR2UZQ5GM7PX44PR6M","target":"record","payload":{"canonical_record":{"source":{"id":"2503.04779","kind":"arxiv","version":4},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.PL","submitted_at":"2025-02-22T13:27:31Z","cross_cats_sorted":["cs.AI","cs.SE"],"title_canon_sha256":"026ae5ab35cfe9c787605c1d4cb9b76ce5518277ca370105283928d16e449a81","abstract_canon_sha256":"19298fead70b53b2494bd617c6b06da35123df2c2e325daf49b08d48ed24e555"},"schema_version":"1.0"},"canonical_sha256":"ca703b238404754cc3a667df7e71f1f333c44750d97876813e7d730edba83e0b","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:11:42.691415Z","signature_b64":"g4b/jZ/L0TQiXl89fQ+PF6PC7fUONdM1//kgF+CxK7x+VzMOK54mrs4SeYM1BKCUxttQ+ZDroXHI0UN7cIloAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ca703b238404754cc3a667df7e71f1f333c44750d97876813e7d730edba83e0b","last_reissued_at":"2026-07-05T11:11:42.690925Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:11:42.690925Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2503.04779","source_version":4,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T11:11:42Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"SSYtx7SpRzMzD3yKMZIEfgvrvGeNOmhIYXk70R/vzR7NlskCYiMg2dxlooAiCJNwWKGPJaTmimvxHnnbNMEbBQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T04:53:03.869813Z"},"content_sha256":"259580199cfa43cb7d9b3cadd24d513684bba33f0f37ccaeabf749775847879a","schema_version":"1.0","event_id":"sha256:259580199cfa43cb7d9b3cadd24d513684bba33f0f37ccaeabf749775847879a"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:ZJYDWI4EAR2UZQ5GM7PX44PR6M","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Can LLMs Reason About Program Semantics? A Comprehensive Evaluation of LLMs on Formal Specification Inference","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.SE"],"primary_cat":"cs.PL","authors_text":"Bach Le, Thanh Le-Cong, Toby Murray","submitted_at":"2025-02-22T13:27:31Z","abstract_excerpt":"Large Language Models (LLMs) are increasingly being used to automate programming tasks. Yet, LLMs' capabilities in reasoning about program semantics are still inadequately studied, leaving significant potential for further exploration. This paper introduces FormalBench, a comprehensive benchmark designed to evaluate LLMs' reasoning abilities on program semantics, particularly via the task of synthesizing formal program specifications to assist verifying program correctness. This task requires both comprehensive reasoning over all possible program executions and the generation of precise, synta"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.04779","kind":"arxiv","version":4},"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/2503.04779/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":"2026-07-05T11:11:42Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Ve1svpQXHWBWkBVTPbvvooWL3EJrI9tmRJRfN5sX2LFYKd8bxF8aPZVcOo50mGHwxdCT6ZuAW2HoH57VAChuAQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T04:53:03.870569Z"},"content_sha256":"36b032ea874b6c850aa1cf8f6511a5c2cb19cbcb69fa3442109246afdf207bb9","schema_version":"1.0","event_id":"sha256:36b032ea874b6c850aa1cf8f6511a5c2cb19cbcb69fa3442109246afdf207bb9"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/ZJYDWI4EAR2UZQ5GM7PX44PR6M/bundle.json","state_url":"https://pith.science/pith/ZJYDWI4EAR2UZQ5GM7PX44PR6M/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/ZJYDWI4EAR2UZQ5GM7PX44PR6M/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":"2026-08-09T04:53:03Z","links":{"resolver":"https://pith.science/pith/ZJYDWI4EAR2UZQ5GM7PX44PR6M","bundle":"https://pith.science/pith/ZJYDWI4EAR2UZQ5GM7PX44PR6M/bundle.json","state":"https://pith.science/pith/ZJYDWI4EAR2UZQ5GM7PX44PR6M/state.json","well_known_bundle":"https://pith.science/.well-known/pith/ZJYDWI4EAR2UZQ5GM7PX44PR6M/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:ZJYDWI4EAR2UZQ5GM7PX44PR6M","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":"19298fead70b53b2494bd617c6b06da35123df2c2e325daf49b08d48ed24e555","cross_cats_sorted":["cs.AI","cs.SE"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.PL","submitted_at":"2025-02-22T13:27:31Z","title_canon_sha256":"026ae5ab35cfe9c787605c1d4cb9b76ce5518277ca370105283928d16e449a81"},"schema_version":"1.0","source":{"id":"2503.04779","kind":"arxiv","version":4}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2503.04779","created_at":"2026-07-05T11:11:42Z"},{"alias_kind":"arxiv_version","alias_value":"2503.04779v4","created_at":"2026-07-05T11:11:42Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.04779","created_at":"2026-07-05T11:11:42Z"},{"alias_kind":"pith_short_12","alias_value":"ZJYDWI4EAR2U","created_at":"2026-07-05T11:11:42Z"},{"alias_kind":"pith_short_16","alias_value":"ZJYDWI4EAR2UZQ5G","created_at":"2026-07-05T11:11:42Z"},{"alias_kind":"pith_short_8","alias_value":"ZJYDWI4E","created_at":"2026-07-05T11:11:42Z"}],"graph_snapshots":[{"event_id":"sha256:36b032ea874b6c850aa1cf8f6511a5c2cb19cbcb69fa3442109246afdf207bb9","target":"graph","created_at":"2026-07-05T11:11:42Z","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/2503.04779/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large Language Models (LLMs) are increasingly being used to automate programming tasks. Yet, LLMs' capabilities in reasoning about program semantics are still inadequately studied, leaving significant potential for further exploration. This paper introduces FormalBench, a comprehensive benchmark designed to evaluate LLMs' reasoning abilities on program semantics, particularly via the task of synthesizing formal program specifications to assist verifying program correctness. This task requires both comprehensive reasoning over all possible program executions and the generation of precise, synta","authors_text":"Bach Le, Thanh Le-Cong, Toby Murray","cross_cats":["cs.AI","cs.SE"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.PL","submitted_at":"2025-02-22T13:27:31Z","title":"Can LLMs Reason About Program Semantics? A Comprehensive Evaluation of LLMs on Formal Specification Inference"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.04779","kind":"arxiv","version":4},"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":"sha256:259580199cfa43cb7d9b3cadd24d513684bba33f0f37ccaeabf749775847879a","target":"record","created_at":"2026-07-05T11:11:42Z","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":"19298fead70b53b2494bd617c6b06da35123df2c2e325daf49b08d48ed24e555","cross_cats_sorted":["cs.AI","cs.SE"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.PL","submitted_at":"2025-02-22T13:27:31Z","title_canon_sha256":"026ae5ab35cfe9c787605c1d4cb9b76ce5518277ca370105283928d16e449a81"},"schema_version":"1.0","source":{"id":"2503.04779","kind":"arxiv","version":4}},"canonical_sha256":"ca703b238404754cc3a667df7e71f1f333c44750d97876813e7d730edba83e0b","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"ca703b238404754cc3a667df7e71f1f333c44750d97876813e7d730edba83e0b","first_computed_at":"2026-07-05T11:11:42.690925Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:11:42.690925Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"g4b/jZ/L0TQiXl89fQ+PF6PC7fUONdM1//kgF+CxK7x+VzMOK54mrs4SeYM1BKCUxttQ+ZDroXHI0UN7cIloAg==","signature_status":"signed_v1","signed_at":"2026-07-05T11:11:42.691415Z","signed_message":"canonical_sha256_bytes"},"source_id":"2503.04779","source_kind":"arxiv","source_version":4}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:259580199cfa43cb7d9b3cadd24d513684bba33f0f37ccaeabf749775847879a","sha256:36b032ea874b6c850aa1cf8f6511a5c2cb19cbcb69fa3442109246afdf207bb9"],"state_sha256":"f3a9a878090e5d52d9c78682ed01db525212e5f283032ed38e90fab3efe75507"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"wzss0isp32A0XcnrDdK+usmLbtz1uU+jrqvT5vn+rlppihJQI56NrmJhC7MkolVes9ti6Pde/kwyDkKGEHePAg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-09T04:53:03.877058Z","bundle_sha256":"b58b3003aac736f890b78a76447fa7e70c944d49403d8f678ddcb841f81f6b2a"}}