{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:XXUS75CFXVARSPLTKEHEN47JGP","short_pith_number":"pith:XXUS75CF","schema_version":"1.0","canonical_sha256":"bde92ff445bd41193d73510e46f3e933df0c0be782acf0ebba5b88b7d9581d3f","source":{"kind":"arxiv","id":"2602.16699","version":3},"attestation_state":"computed","paper":{"title":"Calibrate-Then-Act: Cost-Aware Exploration in LLM Agents","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Greg Durrett, Nicholas Tomlin, Wenxuan Ding","submitted_at":"2026-02-18T18:46:14Z","abstract_excerpt":"LLM agents are deployed in environments where they must interact to acquire information. In these scenarios, the agent must reason about inherent cost-uncertainty tradeoffs in how to act, such as when to stop exploring and commit to an answer. For instance, on a programming task, an agent might run the code it generates, or it might generate tests for that code snippet; the cost of writing and running a test is nonzero, but typically lower than the cost of running buggy code. In this work, we show that we can induce LLM agents to explicitly reason about balancing these cost-uncertainty tradeof"},"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":"2602.16699","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-02-18T18:46:14Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"06681ee244bebf6569380b69fd875f5cef97e3602ebccc0fa2bc2d601940679a","abstract_canon_sha256":"f01015937148712e658bf5622487c6ac3a7cb3f937cc65dcb9bc2d8a1d2e0c72"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T00:02:09.889811Z","signature_b64":"WI8VPlBaIchwIjcpUYH8pIlfgmtjX4eaHhVARQgEqd1SScVGIOdhVZqyL7u3reGl8v3eUdvOIxLx7L3C+zLxCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"bde92ff445bd41193d73510e46f3e933df0c0be782acf0ebba5b88b7d9581d3f","last_reissued_at":"2026-05-20T00:02:09.888944Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T00:02:09.888944Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Calibrate-Then-Act: Cost-Aware Exploration in LLM Agents","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Greg Durrett, Nicholas Tomlin, Wenxuan Ding","submitted_at":"2026-02-18T18:46:14Z","abstract_excerpt":"LLM agents are deployed in environments where they must interact to acquire information. In these scenarios, the agent must reason about inherent cost-uncertainty tradeoffs in how to act, such as when to stop exploring and commit to an answer. For instance, on a programming task, an agent might run the code it generates, or it might generate tests for that code snippet; the cost of writing and running a test is nonzero, but typically lower than the cost of running buggy code. In this work, we show that we can induce LLM agents to explicitly reason about balancing these cost-uncertainty tradeof"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2602.16699","kind":"arxiv","version":3},"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/2602.16699/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":"2602.16699","created_at":"2026-05-20T00:02:09.889087+00:00"},{"alias_kind":"arxiv_version","alias_value":"2602.16699v3","created_at":"2026-05-20T00:02:09.889087+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2602.16699","created_at":"2026-05-20T00:02:09.889087+00:00"},{"alias_kind":"pith_short_12","alias_value":"XXUS75CFXVAR","created_at":"2026-05-20T00:02:09.889087+00:00"},{"alias_kind":"pith_short_16","alias_value":"XXUS75CFXVARSPLT","created_at":"2026-05-20T00:02:09.889087+00:00"},{"alias_kind":"pith_short_8","alias_value":"XXUS75CF","created_at":"2026-05-20T00:02:09.889087+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":2,"sample":[{"citing_arxiv_id":"2604.22577","citing_title":"QuantClaw: Precision Where It Matters for OpenClaw","ref_index":18,"is_internal_anchor":true},{"citing_arxiv_id":"2604.13151","citing_title":"Exploration and Exploitation Errors Are Measurable for Language Model Agents","ref_index":2,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/XXUS75CFXVARSPLTKEHEN47JGP","json":"https://pith.science/pith/XXUS75CFXVARSPLTKEHEN47JGP.json","graph_json":"https://pith.science/api/pith-number/XXUS75CFXVARSPLTKEHEN47JGP/graph.json","events_json":"https://pith.science/api/pith-number/XXUS75CFXVARSPLTKEHEN47JGP/events.json","paper":"https://pith.science/paper/XXUS75CF"},"agent_actions":{"view_html":"https://pith.science/pith/XXUS75CFXVARSPLTKEHEN47JGP","download_json":"https://pith.science/pith/XXUS75CFXVARSPLTKEHEN47JGP.json","view_paper":"https://pith.science/paper/XXUS75CF","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2602.16699&json=true","fetch_graph":"https://pith.science/api/pith-number/XXUS75CFXVARSPLTKEHEN47JGP/graph.json","fetch_events":"https://pith.science/api/pith-number/XXUS75CFXVARSPLTKEHEN47JGP/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/XXUS75CFXVARSPLTKEHEN47JGP/action/timestamp_anchor","attest_storage":"https://pith.science/pith/XXUS75CFXVARSPLTKEHEN47JGP/action/storage_attestation","attest_author":"https://pith.science/pith/XXUS75CFXVARSPLTKEHEN47JGP/action/author_attestation","sign_citation":"https://pith.science/pith/XXUS75CFXVARSPLTKEHEN47JGP/action/citation_signature","submit_replication":"https://pith.science/pith/XXUS75CFXVARSPLTKEHEN47JGP/action/replication_record"}},"created_at":"2026-05-20T00:02:09.889087+00:00","updated_at":"2026-05-20T00:02:09.889087+00:00"}