{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:FCXG5MSUUUPSRJFGFVRX32YZFY","short_pith_number":"pith:FCXG5MSU","schema_version":"1.0","canonical_sha256":"28ae6eb254a51f28a4a62d637deb192e20a1af426a99087f07f557768688e741","source":{"kind":"arxiv","id":"2303.10138","version":1},"attestation_state":"computed","paper":{"title":"Generate, Transform, Answer: Question Specific Tool Synthesis for Tabular Data","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Carlos Gemmell, Jeffrey Dalton","submitted_at":"2023-03-17T17:26:56Z","abstract_excerpt":"Tabular question answering (TQA) presents a challenging setting for neural systems by requiring joint reasoning of natural language with large amounts of semi-structured data. Unlike humans who use programmatic tools like filters to transform data before processing, language models in TQA process tables directly, resulting in information loss as table size increases. In this paper we propose ToolWriter to generate query specific programs and detect when to apply them to transform tables and align them with the TQA model's capabilities. Focusing ToolWriter to generate row-filtering tools improv"},"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":"2303.10138","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2023-03-17T17:26:56Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"81127ab4190c1ae45e89dd1b5fd102302a38fa7ec8be6ff030d5ed49ee35d522","abstract_canon_sha256":"03aeb2f78c9e56da3bc9a950585448e521ebcc3beeb22bea246e69d3feea86ec"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:52:14.692115Z","signature_b64":"g0HLHbuqwJNJshMBZGqCMdOpilEBky9v8LtsrAzPs6zRzE+d4R9RIFrJND/wM4Pctu0xBHxIboLntEZFKhfXBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"28ae6eb254a51f28a4a62d637deb192e20a1af426a99087f07f557768688e741","last_reissued_at":"2026-07-05T05:52:14.691668Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:52:14.691668Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Generate, Transform, Answer: Question Specific Tool Synthesis for Tabular Data","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Carlos Gemmell, Jeffrey Dalton","submitted_at":"2023-03-17T17:26:56Z","abstract_excerpt":"Tabular question answering (TQA) presents a challenging setting for neural systems by requiring joint reasoning of natural language with large amounts of semi-structured data. Unlike humans who use programmatic tools like filters to transform data before processing, language models in TQA process tables directly, resulting in information loss as table size increases. In this paper we propose ToolWriter to generate query specific programs and detect when to apply them to transform tables and align them with the TQA model's capabilities. Focusing ToolWriter to generate row-filtering tools improv"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2303.10138","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/2303.10138/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":"2303.10138","created_at":"2026-07-05T05:52:14.691725+00:00"},{"alias_kind":"arxiv_version","alias_value":"2303.10138v1","created_at":"2026-07-05T05:52:14.691725+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2303.10138","created_at":"2026-07-05T05:52:14.691725+00:00"},{"alias_kind":"pith_short_12","alias_value":"FCXG5MSUUUPS","created_at":"2026-07-05T05:52:14.691725+00:00"},{"alias_kind":"pith_short_16","alias_value":"FCXG5MSUUUPSRJFG","created_at":"2026-07-05T05:52:14.691725+00:00"},{"alias_kind":"pith_short_8","alias_value":"FCXG5MSU","created_at":"2026-07-05T05:52:14.691725+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2507.10281","citing_title":"Toward Real-World Table Agents: Capabilities, Workflows, and Design Principles for LLM-based Table Intelligence","ref_index":2023,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/FCXG5MSUUUPSRJFGFVRX32YZFY","json":"https://pith.science/pith/FCXG5MSUUUPSRJFGFVRX32YZFY.json","graph_json":"https://pith.science/api/pith-number/FCXG5MSUUUPSRJFGFVRX32YZFY/graph.json","events_json":"https://pith.science/api/pith-number/FCXG5MSUUUPSRJFGFVRX32YZFY/events.json","paper":"https://pith.science/paper/FCXG5MSU"},"agent_actions":{"view_html":"https://pith.science/pith/FCXG5MSUUUPSRJFGFVRX32YZFY","download_json":"https://pith.science/pith/FCXG5MSUUUPSRJFGFVRX32YZFY.json","view_paper":"https://pith.science/paper/FCXG5MSU","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2303.10138&json=true","fetch_graph":"https://pith.science/api/pith-number/FCXG5MSUUUPSRJFGFVRX32YZFY/graph.json","fetch_events":"https://pith.science/api/pith-number/FCXG5MSUUUPSRJFGFVRX32YZFY/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/FCXG5MSUUUPSRJFGFVRX32YZFY/action/timestamp_anchor","attest_storage":"https://pith.science/pith/FCXG5MSUUUPSRJFGFVRX32YZFY/action/storage_attestation","attest_author":"https://pith.science/pith/FCXG5MSUUUPSRJFGFVRX32YZFY/action/author_attestation","sign_citation":"https://pith.science/pith/FCXG5MSUUUPSRJFGFVRX32YZFY/action/citation_signature","submit_replication":"https://pith.science/pith/FCXG5MSUUUPSRJFGFVRX32YZFY/action/replication_record"}},"created_at":"2026-07-05T05:52:14.691725+00:00","updated_at":"2026-07-05T05:52:14.691725+00:00"}