{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:L3BSYQDBE4A4CNL5J6WVFIDOFB","short_pith_number":"pith:L3BSYQDB","schema_version":"1.0","canonical_sha256":"5ec32c40612701c1357d4fad52a06e2867b4804d4487719b4c83bc43c40ffadb","source":{"kind":"arxiv","id":"2402.01117","version":1},"attestation_state":"computed","paper":{"title":"DTS-SQL: Decomposed Text-to-SQL with Small Large Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.DB","cs.HC"],"primary_cat":"cs.CL","authors_text":"Davood Rafiei, Mohammadreza Pourreza","submitted_at":"2024-02-02T03:21:00Z","abstract_excerpt":"Leading models for the text-to-SQL task heavily rely on proprietary Large Language Models (LLMs), posing concerns over data privacy. Closing the performance gap between small open-source models and large proprietary models is crucial to mitigate this reliance. To this end, we introduce a novel two-stage fine-tuning approach that decomposes the task into two simpler tasks. Through comprehensive evaluation on two large cross-domain datasets and two small LLMs, we show that this approach improves execution accuracy by 3 to 7 percent, effectively aligning the performance of open-source models with"},"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":"2402.01117","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-02-02T03:21:00Z","cross_cats_sorted":["cs.DB","cs.HC"],"title_canon_sha256":"35c73cbf5f0f7a8332833f50b56c37def94d9fce80b6074442c4ad2b4d2825db","abstract_canon_sha256":"ad157f54f864b897d697ba39f3866692aebcdce58303cd87a394d91bb2e20658"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:40:30.965552Z","signature_b64":"kreOxRMq54Z6A9btz7Rk0+HcbKb/2SdI735zmAhqRVe/KfuY62vVXsZkh1L5wHboINixMi1DIKnTwISr2iwiDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5ec32c40612701c1357d4fad52a06e2867b4804d4487719b4c83bc43c40ffadb","last_reissued_at":"2026-07-05T07:40:30.965083Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:40:30.965083Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"DTS-SQL: Decomposed Text-to-SQL with Small Large Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.DB","cs.HC"],"primary_cat":"cs.CL","authors_text":"Davood Rafiei, Mohammadreza Pourreza","submitted_at":"2024-02-02T03:21:00Z","abstract_excerpt":"Leading models for the text-to-SQL task heavily rely on proprietary Large Language Models (LLMs), posing concerns over data privacy. Closing the performance gap between small open-source models and large proprietary models is crucial to mitigate this reliance. To this end, we introduce a novel two-stage fine-tuning approach that decomposes the task into two simpler tasks. Through comprehensive evaluation on two large cross-domain datasets and two small LLMs, we show that this approach improves execution accuracy by 3 to 7 percent, effectively aligning the performance of open-source models with"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2402.01117","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/2402.01117/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":"2402.01117","created_at":"2026-07-05T07:40:30.965144+00:00"},{"alias_kind":"arxiv_version","alias_value":"2402.01117v1","created_at":"2026-07-05T07:40:30.965144+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2402.01117","created_at":"2026-07-05T07:40:30.965144+00:00"},{"alias_kind":"pith_short_12","alias_value":"L3BSYQDBE4A4","created_at":"2026-07-05T07:40:30.965144+00:00"},{"alias_kind":"pith_short_16","alias_value":"L3BSYQDBE4A4CNL5","created_at":"2026-07-05T07:40:30.965144+00:00"},{"alias_kind":"pith_short_8","alias_value":"L3BSYQDB","created_at":"2026-07-05T07:40:30.965144+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":5,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.05906","citing_title":"ACE-SQL: Adaptive Co-Optimization via Empirical Credit Assignment for Text-to-SQL","ref_index":47,"is_internal_anchor":false},{"citing_arxiv_id":"2502.12911","citing_title":"Knapsack Optimization-based Schema Linking for LLM-based Text-to-SQL Generation","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2405.16755","citing_title":"CHESS: Contextual Harnessing for Efficient SQL Synthesis","ref_index":71,"is_internal_anchor":false},{"citing_arxiv_id":"2507.04701","citing_title":"XiYan-SQL: A Novel Multi-Generator Framework For Text-to-SQL","ref_index":15,"is_internal_anchor":false},{"citing_arxiv_id":"2604.07041","citing_title":"AV-SQL: Decomposing Complex Text-to-SQL Queries with Agentic Views","ref_index":34,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/L3BSYQDBE4A4CNL5J6WVFIDOFB","json":"https://pith.science/pith/L3BSYQDBE4A4CNL5J6WVFIDOFB.json","graph_json":"https://pith.science/api/pith-number/L3BSYQDBE4A4CNL5J6WVFIDOFB/graph.json","events_json":"https://pith.science/api/pith-number/L3BSYQDBE4A4CNL5J6WVFIDOFB/events.json","paper":"https://pith.science/paper/L3BSYQDB"},"agent_actions":{"view_html":"https://pith.science/pith/L3BSYQDBE4A4CNL5J6WVFIDOFB","download_json":"https://pith.science/pith/L3BSYQDBE4A4CNL5J6WVFIDOFB.json","view_paper":"https://pith.science/paper/L3BSYQDB","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2402.01117&json=true","fetch_graph":"https://pith.science/api/pith-number/L3BSYQDBE4A4CNL5J6WVFIDOFB/graph.json","fetch_events":"https://pith.science/api/pith-number/L3BSYQDBE4A4CNL5J6WVFIDOFB/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/L3BSYQDBE4A4CNL5J6WVFIDOFB/action/timestamp_anchor","attest_storage":"https://pith.science/pith/L3BSYQDBE4A4CNL5J6WVFIDOFB/action/storage_attestation","attest_author":"https://pith.science/pith/L3BSYQDBE4A4CNL5J6WVFIDOFB/action/author_attestation","sign_citation":"https://pith.science/pith/L3BSYQDBE4A4CNL5J6WVFIDOFB/action/citation_signature","submit_replication":"https://pith.science/pith/L3BSYQDBE4A4CNL5J6WVFIDOFB/action/replication_record"}},"created_at":"2026-07-05T07:40:30.965144+00:00","updated_at":"2026-07-05T07:40:30.965144+00:00"}