{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:DDRJVLMITEHJI2RBCXRXJQPZHU","short_pith_number":"pith:DDRJVLMI","schema_version":"1.0","canonical_sha256":"18e29aad88990e946a2115e374c1f93d398bf0be540d37d27cf74d7da115f846","source":{"kind":"arxiv","id":"2206.01962","version":2},"attestation_state":"computed","paper":{"title":"Formal Specifications from Natural Language","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","cs.PL"],"primary_cat":"cs.SE","authors_text":"Bernd Finkbeiner, Christopher Hahn, Frederik Schmitt, Julia J. Tillman, Julian Siber, Niklas Metzger","submitted_at":"2022-06-04T10:49:30Z","abstract_excerpt":"We study the generalization abilities of language models when translating natural language into formal specifications with complex semantics. In particular, we fine-tune language models on three datasets consisting of English sentences and their corresponding formal representation: 1) regular expressions (regex), frequently used in programming and search; 2) First-order logic (FOL), commonly used in software verification and theorem proving; and 3) linear-time temporal logic (LTL), which forms the basis for industrial hardware specification languages. Our experiments show that, in these divers"},"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":"2206.01962","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.SE","submitted_at":"2022-06-04T10:49:30Z","cross_cats_sorted":["cs.LG","cs.PL"],"title_canon_sha256":"7a8f2451e83c6f9f5fe6fb4bb8acec1bfcd92647bcc93e6b0ea857105cd61083","abstract_canon_sha256":"8f46020763d735e25d2f067ca4dcf1f3e4f3c6da0fdbb7f05de35977db0448f8"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:08:28.898003Z","signature_b64":"Qs2/zrPNhVHG1c0CyEfsJ7HoFFXtmUppUe0t6EdVEUwBfxwKtQJ/alwqusIdCN79UqPsGRF1wRYt9VFvuArpBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"18e29aad88990e946a2115e374c1f93d398bf0be540d37d27cf74d7da115f846","last_reissued_at":"2026-07-05T05:08:28.897531Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:08:28.897531Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Formal Specifications from Natural Language","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","cs.PL"],"primary_cat":"cs.SE","authors_text":"Bernd Finkbeiner, Christopher Hahn, Frederik Schmitt, Julia J. Tillman, Julian Siber, Niklas Metzger","submitted_at":"2022-06-04T10:49:30Z","abstract_excerpt":"We study the generalization abilities of language models when translating natural language into formal specifications with complex semantics. In particular, we fine-tune language models on three datasets consisting of English sentences and their corresponding formal representation: 1) regular expressions (regex), frequently used in programming and search; 2) First-order logic (FOL), commonly used in software verification and theorem proving; and 3) linear-time temporal logic (LTL), which forms the basis for industrial hardware specification languages. Our experiments show that, in these divers"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2206.01962","kind":"arxiv","version":2},"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/2206.01962/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":"2206.01962","created_at":"2026-07-05T05:08:28.897589+00:00"},{"alias_kind":"arxiv_version","alias_value":"2206.01962v2","created_at":"2026-07-05T05:08:28.897589+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2206.01962","created_at":"2026-07-05T05:08:28.897589+00:00"},{"alias_kind":"pith_short_12","alias_value":"DDRJVLMITEHJ","created_at":"2026-07-05T05:08:28.897589+00:00"},{"alias_kind":"pith_short_16","alias_value":"DDRJVLMITEHJI2RB","created_at":"2026-07-05T05:08:28.897589+00:00"},{"alias_kind":"pith_short_8","alias_value":"DDRJVLMI","created_at":"2026-07-05T05:08:28.897589+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":4,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.05792","citing_title":"Can LLMs Write Correct TLA+ Specifications? Evaluating Natural-Language-to-TLA+ Generation","ref_index":13,"is_internal_anchor":false},{"citing_arxiv_id":"2606.02837","citing_title":"Fixing FOLIO and MALLS: Verified Annotations and an LLM-assisted Framework to Focus Human Relabeling","ref_index":30,"is_internal_anchor":false},{"citing_arxiv_id":"2605.10462","citing_title":"Coherency through formalisations of Structured Natural Language, A case study on FRETish","ref_index":23,"is_internal_anchor":false},{"citing_arxiv_id":"2605.03356","citing_title":"POSTCONDBENCH: Benchmarking Correctness and Completeness in Formal Postcondition Inference","ref_index":126,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/DDRJVLMITEHJI2RBCXRXJQPZHU","json":"https://pith.science/pith/DDRJVLMITEHJI2RBCXRXJQPZHU.json","graph_json":"https://pith.science/api/pith-number/DDRJVLMITEHJI2RBCXRXJQPZHU/graph.json","events_json":"https://pith.science/api/pith-number/DDRJVLMITEHJI2RBCXRXJQPZHU/events.json","paper":"https://pith.science/paper/DDRJVLMI"},"agent_actions":{"view_html":"https://pith.science/pith/DDRJVLMITEHJI2RBCXRXJQPZHU","download_json":"https://pith.science/pith/DDRJVLMITEHJI2RBCXRXJQPZHU.json","view_paper":"https://pith.science/paper/DDRJVLMI","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2206.01962&json=true","fetch_graph":"https://pith.science/api/pith-number/DDRJVLMITEHJI2RBCXRXJQPZHU/graph.json","fetch_events":"https://pith.science/api/pith-number/DDRJVLMITEHJI2RBCXRXJQPZHU/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/DDRJVLMITEHJI2RBCXRXJQPZHU/action/timestamp_anchor","attest_storage":"https://pith.science/pith/DDRJVLMITEHJI2RBCXRXJQPZHU/action/storage_attestation","attest_author":"https://pith.science/pith/DDRJVLMITEHJI2RBCXRXJQPZHU/action/author_attestation","sign_citation":"https://pith.science/pith/DDRJVLMITEHJI2RBCXRXJQPZHU/action/citation_signature","submit_replication":"https://pith.science/pith/DDRJVLMITEHJI2RBCXRXJQPZHU/action/replication_record"}},"created_at":"2026-07-05T05:08:28.897589+00:00","updated_at":"2026-07-05T05:08:28.897589+00:00"}