{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:W4JHEFSEPEIB276RA5KWUOK6N7","short_pith_number":"pith:W4JHEFSE","schema_version":"1.0","canonical_sha256":"b71272164479101d7fd107556a395e6ffa7dc84afafcb07d4c77f2a714ec9848","source":{"kind":"arxiv","id":"2505.00016","version":2},"attestation_state":"computed","paper":{"title":"Sparks of Tabular Reasoning via Text2SQL Reinforcement Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Josefa Lia Stoisser, Julien Fauqueur, Marc Boubnovski Martell","submitted_at":"2025-04-23T19:02:04Z","abstract_excerpt":"This work reframes the Text-to-SQL task as a pathway for teaching large language models (LLMs) to reason over and manipulate tabular data--moving beyond the traditional focus on query generation. We propose a two-stage framework that leverages SQL supervision to develop transferable table reasoning capabilities. First, we synthesize detailed chain-of-thought (CoT) traces from real-world SQL queries, providing step-by-step, clause-level supervision that teaches the model how to traverse, filter, and aggregate table fields. Second, we introduce a Group Relative Policy Optimization (GRPO) reinfor"},"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":"2505.00016","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-04-23T19:02:04Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"5589708a98f58cc189ffc22199afc05496943ea6e5b62081bf2a36171434b0cf","abstract_canon_sha256":"189c5d5d284df58c1a3765aeb773727c3f55b4d5d89925ad1afb7b8ea18cc490"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:57:37.362756Z","signature_b64":"L2oJJi0bHhe05MZAtOf9weor1cvg9w87GfCme1PBxkz2zRq6Njt88CpyioC3IvtaitQfmgHfddFr9Lwm6NiWAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b71272164479101d7fd107556a395e6ffa7dc84afafcb07d4c77f2a714ec9848","last_reissued_at":"2026-07-05T10:57:37.362262Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:57:37.362262Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Sparks of Tabular Reasoning via Text2SQL Reinforcement Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Josefa Lia Stoisser, Julien Fauqueur, Marc Boubnovski Martell","submitted_at":"2025-04-23T19:02:04Z","abstract_excerpt":"This work reframes the Text-to-SQL task as a pathway for teaching large language models (LLMs) to reason over and manipulate tabular data--moving beyond the traditional focus on query generation. We propose a two-stage framework that leverages SQL supervision to develop transferable table reasoning capabilities. First, we synthesize detailed chain-of-thought (CoT) traces from real-world SQL queries, providing step-by-step, clause-level supervision that teaches the model how to traverse, filter, and aggregate table fields. Second, we introduce a Group Relative Policy Optimization (GRPO) reinfor"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.00016","kind":"arxiv","version":2},"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/2505.00016/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":"2505.00016","created_at":"2026-07-05T10:57:37.362341+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.00016v2","created_at":"2026-07-05T10:57:37.362341+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.00016","created_at":"2026-07-05T10:57:37.362341+00:00"},{"alias_kind":"pith_short_12","alias_value":"W4JHEFSEPEIB","created_at":"2026-07-05T10:57:37.362341+00:00"},{"alias_kind":"pith_short_16","alias_value":"W4JHEFSEPEIB276R","created_at":"2026-07-05T10:57:37.362341+00:00"},{"alias_kind":"pith_short_8","alias_value":"W4JHEFSE","created_at":"2026-07-05T10:57:37.362341+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2509.02401","citing_title":"Towards Agents That Know When They Don't Know: Uncertainty as a Control Signal for Structured Reasoning","ref_index":48,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/W4JHEFSEPEIB276RA5KWUOK6N7","json":"https://pith.science/pith/W4JHEFSEPEIB276RA5KWUOK6N7.json","graph_json":"https://pith.science/api/pith-number/W4JHEFSEPEIB276RA5KWUOK6N7/graph.json","events_json":"https://pith.science/api/pith-number/W4JHEFSEPEIB276RA5KWUOK6N7/events.json","paper":"https://pith.science/paper/W4JHEFSE"},"agent_actions":{"view_html":"https://pith.science/pith/W4JHEFSEPEIB276RA5KWUOK6N7","download_json":"https://pith.science/pith/W4JHEFSEPEIB276RA5KWUOK6N7.json","view_paper":"https://pith.science/paper/W4JHEFSE","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.00016&json=true","fetch_graph":"https://pith.science/api/pith-number/W4JHEFSEPEIB276RA5KWUOK6N7/graph.json","fetch_events":"https://pith.science/api/pith-number/W4JHEFSEPEIB276RA5KWUOK6N7/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/W4JHEFSEPEIB276RA5KWUOK6N7/action/timestamp_anchor","attest_storage":"https://pith.science/pith/W4JHEFSEPEIB276RA5KWUOK6N7/action/storage_attestation","attest_author":"https://pith.science/pith/W4JHEFSEPEIB276RA5KWUOK6N7/action/author_attestation","sign_citation":"https://pith.science/pith/W4JHEFSEPEIB276RA5KWUOK6N7/action/citation_signature","submit_replication":"https://pith.science/pith/W4JHEFSEPEIB276RA5KWUOK6N7/action/replication_record"}},"created_at":"2026-07-05T10:57:37.362341+00:00","updated_at":"2026-07-05T10:57:37.362341+00:00"}