{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:6KFPTO53HUPR54DSL2TNMZ3JNJ","short_pith_number":"pith:6KFPTO53","schema_version":"1.0","canonical_sha256":"f28af9bbbb3d1f1ef0725ea6d667696a4bacca4ffd047b037248cfceb38e4688","source":{"kind":"arxiv","id":"2401.12869","version":1},"attestation_state":"computed","paper":{"title":"TroVE: Inducing Verifiable and Efficient Toolboxes for Solving Programmatic Tasks","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Daniel Fried, Graham Neubig, Zhiruo Wang","submitted_at":"2024-01-23T16:03:17Z","abstract_excerpt":"Language models (LMs) can solve tasks such as answering questions about tables or images by writing programs. However, using primitive functions often leads to verbose and error-prone programs, and higher-level functions require expert design. To enable better solutions without human labor, we ask code LMs to curate reusable high-level functions, and use them to write solutions. We present TROVE, a training-free method of inducing a verifiable and efficient toolbox of functions, by generating via using, growing, and periodically trimming the toolbox. On 11 datasets from math, table question an"},"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":"2401.12869","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.AI","submitted_at":"2024-01-23T16:03:17Z","cross_cats_sorted":[],"title_canon_sha256":"228933f2b6489ad57d7dfddf2c0a2f964f54debb2d1c4f0c1b3ecff9b449266c","abstract_canon_sha256":"2fface6703fc5f0c50fe1331430972f6f615b4ec5c24ae5b72ded58e8fe3a97f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:36:37.443153Z","signature_b64":"6aJOcDe6ObRf5acElVJiAk+ArKwznUom49jSHYsXGHevkTKaHjX2Ik+Gxn43MNx8D3kghEgLgIG2g4KbWpYLBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f28af9bbbb3d1f1ef0725ea6d667696a4bacca4ffd047b037248cfceb38e4688","last_reissued_at":"2026-07-05T07:36:37.442690Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:36:37.442690Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"TroVE: Inducing Verifiable and Efficient Toolboxes for Solving Programmatic Tasks","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Daniel Fried, Graham Neubig, Zhiruo Wang","submitted_at":"2024-01-23T16:03:17Z","abstract_excerpt":"Language models (LMs) can solve tasks such as answering questions about tables or images by writing programs. However, using primitive functions often leads to verbose and error-prone programs, and higher-level functions require expert design. To enable better solutions without human labor, we ask code LMs to curate reusable high-level functions, and use them to write solutions. We present TROVE, a training-free method of inducing a verifiable and efficient toolbox of functions, by generating via using, growing, and periodically trimming the toolbox. On 11 datasets from math, table question an"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2401.12869","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/2401.12869/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":"2401.12869","created_at":"2026-07-05T07:36:37.442747+00:00"},{"alias_kind":"arxiv_version","alias_value":"2401.12869v1","created_at":"2026-07-05T07:36:37.442747+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2401.12869","created_at":"2026-07-05T07:36:37.442747+00:00"},{"alias_kind":"pith_short_12","alias_value":"6KFPTO53HUPR","created_at":"2026-07-05T07:36:37.442747+00:00"},{"alias_kind":"pith_short_16","alias_value":"6KFPTO53HUPR54DS","created_at":"2026-07-05T07:36:37.442747+00:00"},{"alias_kind":"pith_short_8","alias_value":"6KFPTO53","created_at":"2026-07-05T07:36:37.442747+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":10,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2607.08010","citing_title":"Tool-Making and Self-Evolving LLM Agents in Low-Latency Systems","ref_index":19,"is_internal_anchor":true},{"citing_arxiv_id":"2606.17929","citing_title":"PreAct: Computer-Using Agents that Get Faster on Repeated Tasks","ref_index":51,"is_internal_anchor":false},{"citing_arxiv_id":"2606.05395","citing_title":"VASO: Formally Verifiable Self-Evolving Skills for Physical AI Agents","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2605.07358","citing_title":"A Comprehensive Survey on Agent Skills: Taxonomy, Techniques, and Applications","ref_index":98,"is_internal_anchor":false},{"citing_arxiv_id":"2412.04272","citing_title":"PoTable: Towards Systematic Thinking via Plan-then-Execute Stage Reasoning on Tables","ref_index":33,"is_internal_anchor":false},{"citing_arxiv_id":"2605.07358","citing_title":"A Comprehensive Survey on Agent Skills: Taxonomy, Techniques, and Applications","ref_index":105,"is_internal_anchor":false},{"citing_arxiv_id":"2605.09423","citing_title":"SimWorld Studio: Automatic Environment Generation with Evolving Coding Agent for Embodied Agent Learning","ref_index":86,"is_internal_anchor":false},{"citing_arxiv_id":"2605.09423","citing_title":"SimWorld Studio: Automatic Environment Generation with Evolving Coding Agent for Embodied Agent Learning","ref_index":86,"is_internal_anchor":false},{"citing_arxiv_id":"2605.05485","citing_title":"ReaComp: Compiling LLM Reasoning into Symbolic Solvers for Efficient Program Synthesis","ref_index":11,"is_internal_anchor":false},{"citing_arxiv_id":"2605.07358","citing_title":"A Comprehensive Survey on Agent Skills: Taxonomy, Techniques, and Applications","ref_index":105,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/6KFPTO53HUPR54DSL2TNMZ3JNJ","json":"https://pith.science/pith/6KFPTO53HUPR54DSL2TNMZ3JNJ.json","graph_json":"https://pith.science/api/pith-number/6KFPTO53HUPR54DSL2TNMZ3JNJ/graph.json","events_json":"https://pith.science/api/pith-number/6KFPTO53HUPR54DSL2TNMZ3JNJ/events.json","paper":"https://pith.science/paper/6KFPTO53"},"agent_actions":{"view_html":"https://pith.science/pith/6KFPTO53HUPR54DSL2TNMZ3JNJ","download_json":"https://pith.science/pith/6KFPTO53HUPR54DSL2TNMZ3JNJ.json","view_paper":"https://pith.science/paper/6KFPTO53","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2401.12869&json=true","fetch_graph":"https://pith.science/api/pith-number/6KFPTO53HUPR54DSL2TNMZ3JNJ/graph.json","fetch_events":"https://pith.science/api/pith-number/6KFPTO53HUPR54DSL2TNMZ3JNJ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/6KFPTO53HUPR54DSL2TNMZ3JNJ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/6KFPTO53HUPR54DSL2TNMZ3JNJ/action/storage_attestation","attest_author":"https://pith.science/pith/6KFPTO53HUPR54DSL2TNMZ3JNJ/action/author_attestation","sign_citation":"https://pith.science/pith/6KFPTO53HUPR54DSL2TNMZ3JNJ/action/citation_signature","submit_replication":"https://pith.science/pith/6KFPTO53HUPR54DSL2TNMZ3JNJ/action/replication_record"}},"created_at":"2026-07-05T07:36:37.442747+00:00","updated_at":"2026-07-05T07:36:37.442747+00:00"}