{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:NBO5S5M2VN2SFCEKOVU52ZXEYJ","short_pith_number":"pith:NBO5S5M2","schema_version":"1.0","canonical_sha256":"685dd9759aab7522888a7569dd66e4c27777124dac2522eeb43ee52847e5dfaa","source":{"kind":"arxiv","id":"2603.20004","version":3},"attestation_state":"computed","paper":{"title":"Human-Level Text-to-SQL via Reinforcement Learning on Verified Data, Without Pipeline Engineering","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.DB","authors_text":"Daniel Kang, Tengjun Jin, Yoojin Choi, Yuxuan Zhu","submitted_at":"2026-03-20T14:49:27Z","abstract_excerpt":"Translating natural language questions to SQL queries (Text-to-SQL) is a long-standing problem in database research. Recent efforts have focused on improving accuracy by building increasingly complex multi-stage large LLM pipelines, layering task decomposition, schema linking, and model-based query selection on top of an LLM. Despite this growing complexity, a substantial gap (>10%) between such systems and human experts persists on benchmarks, suggesting that pipeline engineering alone has hit a ceiling.\n  We show that human-level Text-to-SQL performance is achievable by fine-tuning an LLM us"},"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":"2603.20004","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.DB","submitted_at":"2026-03-20T14:49:27Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"3efd9ce607eed51b6cb51ea00d7d92aa88c087e94b7a7d2e93a1b31d6ff1b63f","abstract_canon_sha256":"ca6a9971ba448fd0976e51dc8789f01560b0a3df93ee340fd1d189a4545bf335"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-11T02:18:57.125900Z","signature_b64":"84Zm7p8+rmFAa7K0atMSsDfnNiQ5W9e9ZALg+nxKoVIBPcm00GtZzbeVeQX53w2+gJMr5mNPW1mnnJmXnL4NBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"685dd9759aab7522888a7569dd66e4c27777124dac2522eeb43ee52847e5dfaa","last_reissued_at":"2026-08-11T02:18:57.118748Z","signature_status":"signed_v1","first_computed_at":"2026-08-11T02:18:57.118748Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Human-Level Text-to-SQL via Reinforcement Learning on Verified Data, Without Pipeline Engineering","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.DB","authors_text":"Daniel Kang, Tengjun Jin, Yoojin Choi, Yuxuan Zhu","submitted_at":"2026-03-20T14:49:27Z","abstract_excerpt":"Translating natural language questions to SQL queries (Text-to-SQL) is a long-standing problem in database research. Recent efforts have focused on improving accuracy by building increasingly complex multi-stage large LLM pipelines, layering task decomposition, schema linking, and model-based query selection on top of an LLM. Despite this growing complexity, a substantial gap (>10%) between such systems and human experts persists on benchmarks, suggesting that pipeline engineering alone has hit a ceiling.\n  We show that human-level Text-to-SQL performance is achievable by fine-tuning an LLM us"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2603.20004","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/2603.20004/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":"2603.20004","created_at":"2026-08-11T02:18:57.120674+00:00"},{"alias_kind":"arxiv_version","alias_value":"2603.20004v3","created_at":"2026-08-11T02:18:57.120674+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2603.20004","created_at":"2026-08-11T02:18:57.120674+00:00"},{"alias_kind":"pith_short_12","alias_value":"NBO5S5M2VN2S","created_at":"2026-08-11T02:18:57.120674+00:00"},{"alias_kind":"pith_short_16","alias_value":"NBO5S5M2VN2SFCEK","created_at":"2026-08-11T02:18:57.120674+00:00"},{"alias_kind":"pith_short_8","alias_value":"NBO5S5M2","created_at":"2026-08-11T02:18:57.120674+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/NBO5S5M2VN2SFCEKOVU52ZXEYJ","json":"https://pith.science/pith/NBO5S5M2VN2SFCEKOVU52ZXEYJ.json","graph_json":"https://pith.science/api/pith-number/NBO5S5M2VN2SFCEKOVU52ZXEYJ/graph.json","events_json":"https://pith.science/api/pith-number/NBO5S5M2VN2SFCEKOVU52ZXEYJ/events.json","paper":"https://pith.science/paper/NBO5S5M2"},"agent_actions":{"view_html":"https://pith.science/pith/NBO5S5M2VN2SFCEKOVU52ZXEYJ","download_json":"https://pith.science/pith/NBO5S5M2VN2SFCEKOVU52ZXEYJ.json","view_paper":"https://pith.science/paper/NBO5S5M2","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2603.20004&json=true","fetch_graph":"https://pith.science/api/pith-number/NBO5S5M2VN2SFCEKOVU52ZXEYJ/graph.json","fetch_events":"https://pith.science/api/pith-number/NBO5S5M2VN2SFCEKOVU52ZXEYJ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/NBO5S5M2VN2SFCEKOVU52ZXEYJ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/NBO5S5M2VN2SFCEKOVU52ZXEYJ/action/storage_attestation","attest_author":"https://pith.science/pith/NBO5S5M2VN2SFCEKOVU52ZXEYJ/action/author_attestation","sign_citation":"https://pith.science/pith/NBO5S5M2VN2SFCEKOVU52ZXEYJ/action/citation_signature","submit_replication":"https://pith.science/pith/NBO5S5M2VN2SFCEKOVU52ZXEYJ/action/replication_record"}},"created_at":"2026-08-11T02:18:57.120674+00:00","updated_at":"2026-08-11T02:18:57.120674+00:00"}