{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:5VUGRL5XAAICLYHQNWUPN2ZQMW","short_pith_number":"pith:5VUGRL5X","schema_version":"1.0","canonical_sha256":"ed6868afb7001025e0f06da8f6eb3065911a42273092588b0cc451c0e68599ca","source":{"kind":"arxiv","id":"2506.09359","version":1},"attestation_state":"computed","paper":{"title":"Taming SQL Complexity: LLM-Based Equivalence Evaluation for Text-to-SQL","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Arash Niknafs, Ashish Basran, Carol Szabo, Qingyun Zeng, Simin Ma","submitted_at":"2025-06-11T03:16:39Z","abstract_excerpt":"The rise of Large Language Models (LLMs) has significantly advanced Text-to-SQL (NL2SQL) systems, yet evaluating the semantic equivalence of generated SQL remains a challenge, especially given ambiguous user queries and multiple valid SQL interpretations. This paper explores using LLMs to assess both semantic and a more practical \"weak\" semantic equivalence. We analyze common patterns of SQL equivalence and inequivalence, discuss challenges in LLM-based evaluation."},"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":"2506.09359","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-06-11T03:16:39Z","cross_cats_sorted":[],"title_canon_sha256":"f4f8a3788084587c103619cbb5783515bc269e1bbbc11fb26d81fe0654be7031","abstract_canon_sha256":"961049387b5a22a23e882b91bfcd7a49f376c6cb93b2e3459374ccabd45bd16d"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:19:35.607604Z","signature_b64":"pDAdM2TQEbuXGhFm2hEi112lysxEaNJKapvxY7KVLvIaC8xYvHbPGxZhYeda2X1ZfuwwGt54PSGHJuPt+qqgDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ed6868afb7001025e0f06da8f6eb3065911a42273092588b0cc451c0e68599ca","last_reissued_at":"2026-07-05T11:19:35.607049Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:19:35.607049Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Taming SQL Complexity: LLM-Based Equivalence Evaluation for Text-to-SQL","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Arash Niknafs, Ashish Basran, Carol Szabo, Qingyun Zeng, Simin Ma","submitted_at":"2025-06-11T03:16:39Z","abstract_excerpt":"The rise of Large Language Models (LLMs) has significantly advanced Text-to-SQL (NL2SQL) systems, yet evaluating the semantic equivalence of generated SQL remains a challenge, especially given ambiguous user queries and multiple valid SQL interpretations. This paper explores using LLMs to assess both semantic and a more practical \"weak\" semantic equivalence. We analyze common patterns of SQL equivalence and inequivalence, discuss challenges in LLM-based evaluation."},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.09359","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/2506.09359/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":"2506.09359","created_at":"2026-07-05T11:19:35.607106+00:00"},{"alias_kind":"arxiv_version","alias_value":"2506.09359v1","created_at":"2026-07-05T11:19:35.607106+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.09359","created_at":"2026-07-05T11:19:35.607106+00:00"},{"alias_kind":"pith_short_12","alias_value":"5VUGRL5XAAIC","created_at":"2026-07-05T11:19:35.607106+00:00"},{"alias_kind":"pith_short_16","alias_value":"5VUGRL5XAAICLYHQ","created_at":"2026-07-05T11:19:35.607106+00:00"},{"alias_kind":"pith_short_8","alias_value":"5VUGRL5X","created_at":"2026-07-05T11:19:35.607106+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/5VUGRL5XAAICLYHQNWUPN2ZQMW","json":"https://pith.science/pith/5VUGRL5XAAICLYHQNWUPN2ZQMW.json","graph_json":"https://pith.science/api/pith-number/5VUGRL5XAAICLYHQNWUPN2ZQMW/graph.json","events_json":"https://pith.science/api/pith-number/5VUGRL5XAAICLYHQNWUPN2ZQMW/events.json","paper":"https://pith.science/paper/5VUGRL5X"},"agent_actions":{"view_html":"https://pith.science/pith/5VUGRL5XAAICLYHQNWUPN2ZQMW","download_json":"https://pith.science/pith/5VUGRL5XAAICLYHQNWUPN2ZQMW.json","view_paper":"https://pith.science/paper/5VUGRL5X","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2506.09359&json=true","fetch_graph":"https://pith.science/api/pith-number/5VUGRL5XAAICLYHQNWUPN2ZQMW/graph.json","fetch_events":"https://pith.science/api/pith-number/5VUGRL5XAAICLYHQNWUPN2ZQMW/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/5VUGRL5XAAICLYHQNWUPN2ZQMW/action/timestamp_anchor","attest_storage":"https://pith.science/pith/5VUGRL5XAAICLYHQNWUPN2ZQMW/action/storage_attestation","attest_author":"https://pith.science/pith/5VUGRL5XAAICLYHQNWUPN2ZQMW/action/author_attestation","sign_citation":"https://pith.science/pith/5VUGRL5XAAICLYHQNWUPN2ZQMW/action/citation_signature","submit_replication":"https://pith.science/pith/5VUGRL5XAAICLYHQNWUPN2ZQMW/action/replication_record"}},"created_at":"2026-07-05T11:19:35.607106+00:00","updated_at":"2026-07-05T11:19:35.607106+00:00"}