{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:PDF7C6XI45MHBLJYTQP5GEN4JH","short_pith_number":"pith:PDF7C6XI","schema_version":"1.0","canonical_sha256":"78cbf17ae8e75870ad389c1fd311bc49cf21643f08c9fcd5c3a8392739234753","source":{"kind":"arxiv","id":"2511.12784","version":3},"attestation_state":"computed","paper":{"title":"Evaluating Autoformalization Robustness via Semantically Similar Paraphrasing","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LO"],"primary_cat":"cs.CL","authors_text":"Asfahan Shah, Hayden Moore","submitted_at":"2025-11-16T21:25:59Z","abstract_excerpt":"Large Language Models (LLMs) have recently emerged as powerful tools for autoformalization. Despite their impressive performance, these models can still struggle to produce grounded and verifiable formalizations. Recent work in text-to-SQL, has revealed that LLMs can be sensitive to paraphrased natural language (NL) inputs, even when high degrees of semantic fidelity are preserved. In this paper, we investigate this claim in the autoformalization domain. Specifically, we evaluate the robustness of LLMs generating formal proofs with semantically similar paraphrased NL statements by measuring se"},"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":"2511.12784","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-11-16T21:25:59Z","cross_cats_sorted":["cs.LO"],"title_canon_sha256":"c15a1ec6e04dffc5bdd369f86d4920412c2f010a1dafc8e64eafe25b50b9fe37","abstract_canon_sha256":"4bdd7570e025e03c04da3dc89504e9a902541c577face19c55c9f6add2c82793"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-04T01:09:39.610963Z","signature_b64":"EbhEyfBi/iHboJw2rhx0nf/8+oNO7ARSMBei5awN15XWIREm6mbiTg1Hk2O+9E9PwmnI1gvVfPb5vd/zDxomBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"78cbf17ae8e75870ad389c1fd311bc49cf21643f08c9fcd5c3a8392739234753","last_reissued_at":"2026-06-04T01:09:39.610384Z","signature_status":"signed_v1","first_computed_at":"2026-06-04T01:09:39.610384Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Evaluating Autoformalization Robustness via Semantically Similar Paraphrasing","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LO"],"primary_cat":"cs.CL","authors_text":"Asfahan Shah, Hayden Moore","submitted_at":"2025-11-16T21:25:59Z","abstract_excerpt":"Large Language Models (LLMs) have recently emerged as powerful tools for autoformalization. Despite their impressive performance, these models can still struggle to produce grounded and verifiable formalizations. Recent work in text-to-SQL, has revealed that LLMs can be sensitive to paraphrased natural language (NL) inputs, even when high degrees of semantic fidelity are preserved. In this paper, we investigate this claim in the autoformalization domain. Specifically, we evaluate the robustness of LLMs generating formal proofs with semantically similar paraphrased NL statements by measuring se"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2511.12784","kind":"arxiv","version":3},"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/2511.12784/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":"2511.12784","created_at":"2026-06-04T01:09:39.610454+00:00"},{"alias_kind":"arxiv_version","alias_value":"2511.12784v3","created_at":"2026-06-04T01:09:39.610454+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2511.12784","created_at":"2026-06-04T01:09:39.610454+00:00"},{"alias_kind":"pith_short_12","alias_value":"PDF7C6XI45MH","created_at":"2026-06-04T01:09:39.610454+00:00"},{"alias_kind":"pith_short_16","alias_value":"PDF7C6XI45MHBLJY","created_at":"2026-06-04T01:09:39.610454+00:00"},{"alias_kind":"pith_short_8","alias_value":"PDF7C6XI","created_at":"2026-06-04T01:09:39.610454+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":3,"sample":[{"citing_arxiv_id":"2604.23135","citing_title":"Characterizing Paraphrase-Induced Failures in Lean 4 Autoformalization","ref_index":7,"is_internal_anchor":true},{"citing_arxiv_id":"2605.17255","citing_title":"CAM-Bench: A Benchmark for Computational and Applied Mathematics in Lean","ref_index":25,"is_internal_anchor":true},{"citing_arxiv_id":"2604.23135","citing_title":"Characterizing Paraphrase-Induced Failures in Lean 4 Autoformalization","ref_index":7,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/PDF7C6XI45MHBLJYTQP5GEN4JH","json":"https://pith.science/pith/PDF7C6XI45MHBLJYTQP5GEN4JH.json","graph_json":"https://pith.science/api/pith-number/PDF7C6XI45MHBLJYTQP5GEN4JH/graph.json","events_json":"https://pith.science/api/pith-number/PDF7C6XI45MHBLJYTQP5GEN4JH/events.json","paper":"https://pith.science/paper/PDF7C6XI"},"agent_actions":{"view_html":"https://pith.science/pith/PDF7C6XI45MHBLJYTQP5GEN4JH","download_json":"https://pith.science/pith/PDF7C6XI45MHBLJYTQP5GEN4JH.json","view_paper":"https://pith.science/paper/PDF7C6XI","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2511.12784&json=true","fetch_graph":"https://pith.science/api/pith-number/PDF7C6XI45MHBLJYTQP5GEN4JH/graph.json","fetch_events":"https://pith.science/api/pith-number/PDF7C6XI45MHBLJYTQP5GEN4JH/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/PDF7C6XI45MHBLJYTQP5GEN4JH/action/timestamp_anchor","attest_storage":"https://pith.science/pith/PDF7C6XI45MHBLJYTQP5GEN4JH/action/storage_attestation","attest_author":"https://pith.science/pith/PDF7C6XI45MHBLJYTQP5GEN4JH/action/author_attestation","sign_citation":"https://pith.science/pith/PDF7C6XI45MHBLJYTQP5GEN4JH/action/citation_signature","submit_replication":"https://pith.science/pith/PDF7C6XI45MHBLJYTQP5GEN4JH/action/replication_record"}},"created_at":"2026-06-04T01:09:39.610454+00:00","updated_at":"2026-06-04T01:09:39.610454+00:00"}