{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:3VJECLDBJHHTDQXYAKKW55FCVQ","short_pith_number":"pith:3VJECLDB","schema_version":"1.0","canonical_sha256":"dd52412c6149cf31c2f802956ef4a2ac36f668a39d7d9f11bb0c84e9ce7b3831","source":{"kind":"arxiv","id":"2410.12004","version":1},"attestation_state":"computed","paper":{"title":"Toolken+: Improving LLM Tool Usage with Reranking and a Reject Option","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Andrey Bout, Konstantin Yakovlev, Sergey Nikolenko","submitted_at":"2024-10-15T19:09:03Z","abstract_excerpt":"The recently proposed ToolkenGPT tool learning paradigm demonstrates promising performance but suffers from two major issues: first, it cannot benefit from tool documentation, and second, it often makes mistakes in whether to use a tool at all. We introduce Toolken+ that mitigates the first problem by reranking top $k$ tools selected by ToolkenGPT and the second problem with a special \"Reject\" option such that the model will generate a vocabulary token if \"Reject\" is ranked first. We demonstrate the effectiveness of Toolken+ on multistep numerical reasoning and tool selection tasks."},"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":"2410.12004","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-10-15T19:09:03Z","cross_cats_sorted":[],"title_canon_sha256":"6d6bd40bdce051e80702dd441a05ac9f784d7d9fd42ac2d9c4f134261d672104","abstract_canon_sha256":"ecc7635bf67fea9d1b320f352dac64170e68cda2b5302ce5faa1e958f7f54ca5"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:21:11.536631Z","signature_b64":"cOdYgVaumHhTKBbPR2sj/iEpmgFkNIkTR+e4qFJiKEcT43icAr5t9XT8RINXuCRtGYYBr59sAqmpJLQQHuBoCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"dd52412c6149cf31c2f802956ef4a2ac36f668a39d7d9f11bb0c84e9ce7b3831","last_reissued_at":"2026-07-05T09:21:11.536088Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:21:11.536088Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Toolken+: Improving LLM Tool Usage with Reranking and a Reject Option","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Andrey Bout, Konstantin Yakovlev, Sergey Nikolenko","submitted_at":"2024-10-15T19:09:03Z","abstract_excerpt":"The recently proposed ToolkenGPT tool learning paradigm demonstrates promising performance but suffers from two major issues: first, it cannot benefit from tool documentation, and second, it often makes mistakes in whether to use a tool at all. We introduce Toolken+ that mitigates the first problem by reranking top $k$ tools selected by ToolkenGPT and the second problem with a special \"Reject\" option such that the model will generate a vocabulary token if \"Reject\" is ranked first. We demonstrate the effectiveness of Toolken+ on multistep numerical reasoning and tool selection tasks."},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.12004","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/2410.12004/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":"2410.12004","created_at":"2026-07-05T09:21:11.536154+00:00"},{"alias_kind":"arxiv_version","alias_value":"2410.12004v1","created_at":"2026-07-05T09:21:11.536154+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.12004","created_at":"2026-07-05T09:21:11.536154+00:00"},{"alias_kind":"pith_short_12","alias_value":"3VJECLDBJHHT","created_at":"2026-07-05T09:21:11.536154+00:00"},{"alias_kind":"pith_short_16","alias_value":"3VJECLDBJHHTDQXY","created_at":"2026-07-05T09:21:11.536154+00:00"},{"alias_kind":"pith_short_8","alias_value":"3VJECLDB","created_at":"2026-07-05T09:21:11.536154+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2503.23278","citing_title":"Model Context Protocol (MCP): Landscape, Security Threats, and Future Research Directions","ref_index":74,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/3VJECLDBJHHTDQXYAKKW55FCVQ","json":"https://pith.science/pith/3VJECLDBJHHTDQXYAKKW55FCVQ.json","graph_json":"https://pith.science/api/pith-number/3VJECLDBJHHTDQXYAKKW55FCVQ/graph.json","events_json":"https://pith.science/api/pith-number/3VJECLDBJHHTDQXYAKKW55FCVQ/events.json","paper":"https://pith.science/paper/3VJECLDB"},"agent_actions":{"view_html":"https://pith.science/pith/3VJECLDBJHHTDQXYAKKW55FCVQ","download_json":"https://pith.science/pith/3VJECLDBJHHTDQXYAKKW55FCVQ.json","view_paper":"https://pith.science/paper/3VJECLDB","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2410.12004&json=true","fetch_graph":"https://pith.science/api/pith-number/3VJECLDBJHHTDQXYAKKW55FCVQ/graph.json","fetch_events":"https://pith.science/api/pith-number/3VJECLDBJHHTDQXYAKKW55FCVQ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/3VJECLDBJHHTDQXYAKKW55FCVQ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/3VJECLDBJHHTDQXYAKKW55FCVQ/action/storage_attestation","attest_author":"https://pith.science/pith/3VJECLDBJHHTDQXYAKKW55FCVQ/action/author_attestation","sign_citation":"https://pith.science/pith/3VJECLDBJHHTDQXYAKKW55FCVQ/action/citation_signature","submit_replication":"https://pith.science/pith/3VJECLDBJHHTDQXYAKKW55FCVQ/action/replication_record"}},"created_at":"2026-07-05T09:21:11.536154+00:00","updated_at":"2026-07-05T09:21:11.536154+00:00"}