{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:EMJUC5WJXHOAFBNXOK6LNJOYHR","short_pith_number":"pith:EMJUC5WJ","schema_version":"1.0","canonical_sha256":"23134176c9b9dc0285b772bcb6a5d83c6917cb6e5566f68d7bba8bde89c4a969","source":{"kind":"arxiv","id":"2502.11656","version":1},"attestation_state":"computed","paper":{"title":"Uncovering the Impact of Chain-of-Thought Reasoning for Direct Preference Optimization: Lessons from Text-to-SQL","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.DB"],"primary_cat":"cs.CL","authors_text":"Haiyong Xu, Hanbing Liu, Haoyang Li, Jing Zhang, Qi Qi, Ruotong Chen, Tian Tian, Xiaokang Zhang","submitted_at":"2025-02-17T10:47:17Z","abstract_excerpt":"Direct Preference Optimization (DPO) has proven effective in complex reasoning tasks like math word problems and code generation. However, when applied to Text-to-SQL datasets, it often fails to improve performance and can even degrade it. Our investigation reveals the root cause: unlike math and code tasks, which naturally integrate Chain-of-Thought (CoT) reasoning with DPO, Text-to-SQL datasets typically include only final answers (gold SQL queries) without detailed CoT solutions. By augmenting Text-to-SQL datasets with synthetic CoT solutions, we achieve, for the first time, consistent and "},"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":"2502.11656","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-02-17T10:47:17Z","cross_cats_sorted":["cs.DB"],"title_canon_sha256":"5b5ac3a80fbb3c1eee057582728b3384ae1a7c1e8d596c04bb8b0d68eb5b7572","abstract_canon_sha256":"e65d1ea8008c7d7e9733604af41f4a126fe101a579f06698f917f0212da3a4bb"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:15:32.141900Z","signature_b64":"AHuZn/m9za3fM2qZxZB4EElI7ssZxzzmhShATYM06LUAwAd8u0x+25wloXIOm6Ht0lqA5YEtZ9GetCIlAqRbAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"23134176c9b9dc0285b772bcb6a5d83c6917cb6e5566f68d7bba8bde89c4a969","last_reissued_at":"2026-07-05T10:15:32.141389Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:15:32.141389Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Uncovering the Impact of Chain-of-Thought Reasoning for Direct Preference Optimization: Lessons from Text-to-SQL","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.DB"],"primary_cat":"cs.CL","authors_text":"Haiyong Xu, Hanbing Liu, Haoyang Li, Jing Zhang, Qi Qi, Ruotong Chen, Tian Tian, Xiaokang Zhang","submitted_at":"2025-02-17T10:47:17Z","abstract_excerpt":"Direct Preference Optimization (DPO) has proven effective in complex reasoning tasks like math word problems and code generation. However, when applied to Text-to-SQL datasets, it often fails to improve performance and can even degrade it. Our investigation reveals the root cause: unlike math and code tasks, which naturally integrate Chain-of-Thought (CoT) reasoning with DPO, Text-to-SQL datasets typically include only final answers (gold SQL queries) without detailed CoT solutions. By augmenting Text-to-SQL datasets with synthetic CoT solutions, we achieve, for the first time, consistent and "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.11656","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/2502.11656/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":"2502.11656","created_at":"2026-07-05T10:15:32.141444+00:00"},{"alias_kind":"arxiv_version","alias_value":"2502.11656v1","created_at":"2026-07-05T10:15:32.141444+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.11656","created_at":"2026-07-05T10:15:32.141444+00:00"},{"alias_kind":"pith_short_12","alias_value":"EMJUC5WJXHOA","created_at":"2026-07-05T10:15:32.141444+00:00"},{"alias_kind":"pith_short_16","alias_value":"EMJUC5WJXHOAFBNX","created_at":"2026-07-05T10:15:32.141444+00:00"},{"alias_kind":"pith_short_8","alias_value":"EMJUC5WJ","created_at":"2026-07-05T10:15:32.141444+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.04066","citing_title":"Adapt to Thrive! Adaptive Power-Mean Policy Optimization for Improved LLM Reasoning","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2605.04065","citing_title":"Free Energy-Driven Reinforcement Learning with Adaptive Advantage Shaping for Unsupervised Reasoning in LLMs","ref_index":6,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/EMJUC5WJXHOAFBNXOK6LNJOYHR","json":"https://pith.science/pith/EMJUC5WJXHOAFBNXOK6LNJOYHR.json","graph_json":"https://pith.science/api/pith-number/EMJUC5WJXHOAFBNXOK6LNJOYHR/graph.json","events_json":"https://pith.science/api/pith-number/EMJUC5WJXHOAFBNXOK6LNJOYHR/events.json","paper":"https://pith.science/paper/EMJUC5WJ"},"agent_actions":{"view_html":"https://pith.science/pith/EMJUC5WJXHOAFBNXOK6LNJOYHR","download_json":"https://pith.science/pith/EMJUC5WJXHOAFBNXOK6LNJOYHR.json","view_paper":"https://pith.science/paper/EMJUC5WJ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2502.11656&json=true","fetch_graph":"https://pith.science/api/pith-number/EMJUC5WJXHOAFBNXOK6LNJOYHR/graph.json","fetch_events":"https://pith.science/api/pith-number/EMJUC5WJXHOAFBNXOK6LNJOYHR/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/EMJUC5WJXHOAFBNXOK6LNJOYHR/action/timestamp_anchor","attest_storage":"https://pith.science/pith/EMJUC5WJXHOAFBNXOK6LNJOYHR/action/storage_attestation","attest_author":"https://pith.science/pith/EMJUC5WJXHOAFBNXOK6LNJOYHR/action/author_attestation","sign_citation":"https://pith.science/pith/EMJUC5WJXHOAFBNXOK6LNJOYHR/action/citation_signature","submit_replication":"https://pith.science/pith/EMJUC5WJXHOAFBNXOK6LNJOYHR/action/replication_record"}},"created_at":"2026-07-05T10:15:32.141444+00:00","updated_at":"2026-07-05T10:15:32.141444+00:00"}