{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:BCBKPMDXQZVPMBT3IWTBRLTS4J","short_pith_number":"pith:BCBKPMDX","schema_version":"1.0","canonical_sha256":"0882a7b077866af6067b45a618ae72e2692b06515a61f63af685c9d849490595","source":{"kind":"arxiv","id":"2504.18851","version":1},"attestation_state":"computed","paper":{"title":"When2Call: When (not) to Call Tools","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Ameya Sunil Mahabaleshwarkar, Hayley Ross, Yoshi Suhara","submitted_at":"2025-04-26T08:30:02Z","abstract_excerpt":"Leveraging external tools is a key feature for modern Language Models (LMs) to expand their capabilities and integrate them into existing systems. However, existing benchmarks primarily focus on the accuracy of tool calling -- whether the correct tool is called with the correct parameters -- and less on evaluating when LMs should (not) call tools. We develop a new benchmark, When2Call, which evaluates tool-calling decision-making: when to generate a tool call, when to ask follow-up questions and when to admit the question can't be answered with the tools provided. We find that state-of-the-art"},"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":"2504.18851","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-04-26T08:30:02Z","cross_cats_sorted":[],"title_canon_sha256":"27cf471d9e17d99e920fd332c030701d4cc4fdc32f6130edd3b496e13654f712","abstract_canon_sha256":"74eff34debf39b25160b52ce29f00f8b129446e3e38fb565100d4cb08b43f581"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:54:22.445439Z","signature_b64":"v7xNBUBSVoQjV+DjdzE/s5MLXIDXA9zj9GeKRLbKbYgrfYG67khgvlLOJhVhZJbBOwa3ppD6DLOeGxEnEHJ7AA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0882a7b077866af6067b45a618ae72e2692b06515a61f63af685c9d849490595","last_reissued_at":"2026-07-05T10:54:22.444924Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:54:22.444924Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"When2Call: When (not) to Call Tools","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Ameya Sunil Mahabaleshwarkar, Hayley Ross, Yoshi Suhara","submitted_at":"2025-04-26T08:30:02Z","abstract_excerpt":"Leveraging external tools is a key feature for modern Language Models (LMs) to expand their capabilities and integrate them into existing systems. However, existing benchmarks primarily focus on the accuracy of tool calling -- whether the correct tool is called with the correct parameters -- and less on evaluating when LMs should (not) call tools. We develop a new benchmark, When2Call, which evaluates tool-calling decision-making: when to generate a tool call, when to ask follow-up questions and when to admit the question can't be answered with the tools provided. We find that state-of-the-art"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2504.18851","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/2504.18851/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":"2504.18851","created_at":"2026-07-05T10:54:22.444983+00:00"},{"alias_kind":"arxiv_version","alias_value":"2504.18851v1","created_at":"2026-07-05T10:54:22.444983+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2504.18851","created_at":"2026-07-05T10:54:22.444983+00:00"},{"alias_kind":"pith_short_12","alias_value":"BCBKPMDXQZVP","created_at":"2026-07-05T10:54:22.444983+00:00"},{"alias_kind":"pith_short_16","alias_value":"BCBKPMDXQZVPMBT3","created_at":"2026-07-05T10:54:22.444983+00:00"},{"alias_kind":"pith_short_8","alias_value":"BCBKPMDX","created_at":"2026-07-05T10:54:22.444983+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":4,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.12587","citing_title":"Strategic Decision Support for AI Agents","ref_index":73,"is_internal_anchor":false},{"citing_arxiv_id":"2605.14038","citing_title":"Model-Adaptive Tool Necessity Reveals the Knowing-Doing Gap in LLM Tool Use","ref_index":25,"is_internal_anchor":false},{"citing_arxiv_id":"2510.23853","citing_title":"Your LLM Agents are Temporally Blind: The Misalignment Between Tool Use Decisions and Human Time Perception","ref_index":30,"is_internal_anchor":false},{"citing_arxiv_id":"2605.14038","citing_title":"Model-Adaptive Tool Necessity Reveals the Knowing-Doing Gap in LLM Tool Use","ref_index":25,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/BCBKPMDXQZVPMBT3IWTBRLTS4J","json":"https://pith.science/pith/BCBKPMDXQZVPMBT3IWTBRLTS4J.json","graph_json":"https://pith.science/api/pith-number/BCBKPMDXQZVPMBT3IWTBRLTS4J/graph.json","events_json":"https://pith.science/api/pith-number/BCBKPMDXQZVPMBT3IWTBRLTS4J/events.json","paper":"https://pith.science/paper/BCBKPMDX"},"agent_actions":{"view_html":"https://pith.science/pith/BCBKPMDXQZVPMBT3IWTBRLTS4J","download_json":"https://pith.science/pith/BCBKPMDXQZVPMBT3IWTBRLTS4J.json","view_paper":"https://pith.science/paper/BCBKPMDX","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2504.18851&json=true","fetch_graph":"https://pith.science/api/pith-number/BCBKPMDXQZVPMBT3IWTBRLTS4J/graph.json","fetch_events":"https://pith.science/api/pith-number/BCBKPMDXQZVPMBT3IWTBRLTS4J/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/BCBKPMDXQZVPMBT3IWTBRLTS4J/action/timestamp_anchor","attest_storage":"https://pith.science/pith/BCBKPMDXQZVPMBT3IWTBRLTS4J/action/storage_attestation","attest_author":"https://pith.science/pith/BCBKPMDXQZVPMBT3IWTBRLTS4J/action/author_attestation","sign_citation":"https://pith.science/pith/BCBKPMDXQZVPMBT3IWTBRLTS4J/action/citation_signature","submit_replication":"https://pith.science/pith/BCBKPMDXQZVPMBT3IWTBRLTS4J/action/replication_record"}},"created_at":"2026-07-05T10:54:22.444983+00:00","updated_at":"2026-07-05T10:54:22.444983+00:00"}