{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:C6AL6HGQAENSCXXKUACOMFW25E","short_pith_number":"pith:C6AL6HGQ","schema_version":"1.0","canonical_sha256":"1780bf1cd0011b215eeaa004e616dae93eabb223847f30499bd76f843402d197","source":{"kind":"arxiv","id":"2504.14870","version":2},"attestation_state":"computed","paper":{"title":"Acting Less is Reasoning More! Teaching Model to Act Efficiently","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.AI","authors_text":"Bowen Jin, Cheng Qian, Heng Ji, Hongru Wang, Jiahao Qiu, Kam-Fai Wong, Mengdi Wang, Shijue Huang, Wanjun Zhong, Xiusi Chen","submitted_at":"2025-04-21T05:40:05Z","abstract_excerpt":"Tool-integrated reasoning (TIR) augments large language models (LLMs) with the ability to invoke external tools during long-form reasoning, such as search engines and code interpreters, to solve tasks beyond the capabilities of internal reasoning. While reinforcement learning (RL) has shown promise in training such agents, most of existing approaches typically optimize only for final correctness without considering the efficiency or necessity of external tool use. This often leads to excessive tool calling, incurring high computational costs and hindering the development of internal reasoning "},"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":"2504.14870","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.AI","submitted_at":"2025-04-21T05:40:05Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"e929f5d17b76069d642523814a7dade7948b6ebd9c3b0a78847eda6396d1b237","abstract_canon_sha256":"683b40aead2a3b52d226d8e146bdb9a197be61dc720052c039197e4cb377e2ec"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:13:24.325697Z","signature_b64":"ovEMClPD2hfFPHqCTnTts/p6n0gb242k6m3WPKLEeR9DEvXIpuY1U9HNPIPMwvTLoLKnN3uTAYW8BK2RXEbzBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"1780bf1cd0011b215eeaa004e616dae93eabb223847f30499bd76f843402d197","last_reissued_at":"2026-07-05T11:13:24.325167Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:13:24.325167Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Acting Less is Reasoning More! Teaching Model to Act Efficiently","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.AI","authors_text":"Bowen Jin, Cheng Qian, Heng Ji, Hongru Wang, Jiahao Qiu, Kam-Fai Wong, Mengdi Wang, Shijue Huang, Wanjun Zhong, Xiusi Chen","submitted_at":"2025-04-21T05:40:05Z","abstract_excerpt":"Tool-integrated reasoning (TIR) augments large language models (LLMs) with the ability to invoke external tools during long-form reasoning, such as search engines and code interpreters, to solve tasks beyond the capabilities of internal reasoning. While reinforcement learning (RL) has shown promise in training such agents, most of existing approaches typically optimize only for final correctness without considering the efficiency or necessity of external tool use. This often leads to excessive tool calling, incurring high computational costs and hindering the development of internal reasoning "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2504.14870","kind":"arxiv","version":2},"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/2504.14870/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":"2504.14870","created_at":"2026-07-05T11:13:24.325223+00:00"},{"alias_kind":"arxiv_version","alias_value":"2504.14870v2","created_at":"2026-07-05T11:13:24.325223+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2504.14870","created_at":"2026-07-05T11:13:24.325223+00:00"},{"alias_kind":"pith_short_12","alias_value":"C6AL6HGQAENS","created_at":"2026-07-05T11:13:24.325223+00:00"},{"alias_kind":"pith_short_16","alias_value":"C6AL6HGQAENSCXXK","created_at":"2026-07-05T11:13:24.325223+00:00"},{"alias_kind":"pith_short_8","alias_value":"C6AL6HGQ","created_at":"2026-07-05T11:13:24.325223+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":16,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.26969","citing_title":"Einstein World Models","ref_index":115,"is_internal_anchor":false},{"citing_arxiv_id":"2606.11652","citing_title":"IAPO: Input Attribution-Aware Policy Optimization for Tool Use in Small Multimodal Agents","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2606.06667","citing_title":"The Piggyback Hypothesis of Generalization: Explaining and Mitigating Emergent Misalignment","ref_index":25,"is_internal_anchor":false},{"citing_arxiv_id":"2605.28774","citing_title":"Agent Explorative Policy Optimization for Multimodal Agentic Reasoning","ref_index":33,"is_internal_anchor":false},{"citing_arxiv_id":"2605.29894","citing_title":"Train the Agent, Not the Expert: Learning to Harness Heterogeneous Experts for Multi-Turn Visual Reasoning","ref_index":29,"is_internal_anchor":false},{"citing_arxiv_id":"2509.26383","citing_title":"Efficient and Transferable Agentic Knowledge Graph RAG via Reinforcement Learning","ref_index":38,"is_internal_anchor":false},{"citing_arxiv_id":"2605.20743","citing_title":"Draw2Think: Harnessing Geometry Reasoning through Constraint Engine Interaction","ref_index":49,"is_internal_anchor":false},{"citing_arxiv_id":"2509.02547","citing_title":"The Landscape of Agentic Reinforcement Learning for LLMs: A Survey","ref_index":101,"is_internal_anchor":false},{"citing_arxiv_id":"2510.01152","citing_title":"MASH: Modeling Abstention via Selective Help-Seeking","ref_index":17,"is_internal_anchor":false},{"citing_arxiv_id":"2504.21776","citing_title":"WebThinker: Empowering Large Reasoning Models with Deep Research Capability","ref_index":50,"is_internal_anchor":false},{"citing_arxiv_id":"2509.02544","citing_title":"UI-TARS-2 Technical Report: Advancing GUI Agent with Multi-Turn Reinforcement Learning","ref_index":70,"is_internal_anchor":false},{"citing_arxiv_id":"2605.12481","citing_title":"ToolCUA: Towards Optimal GUI-Tool Path Orchestration for Computer Use Agents","ref_index":38,"is_internal_anchor":false},{"citing_arxiv_id":"2605.09544","citing_title":"TIDE-Bench: Task-Aware and Diagnostic Evaluation of Tool-Integrated Reasoning","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2505.10978","citing_title":"Group-in-Group Policy Optimization for LLM Agent Training","ref_index":61,"is_internal_anchor":false},{"citing_arxiv_id":"2604.09455","citing_title":"E3-TIR: Enhanced Experience Exploitation for Tool-Integrated Reasoning","ref_index":28,"is_internal_anchor":false},{"citing_arxiv_id":"2604.18292","citing_title":"Agent-World: Scaling Real-World Environment Synthesis for Evolving General Agent Intelligence","ref_index":100,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/C6AL6HGQAENSCXXKUACOMFW25E","json":"https://pith.science/pith/C6AL6HGQAENSCXXKUACOMFW25E.json","graph_json":"https://pith.science/api/pith-number/C6AL6HGQAENSCXXKUACOMFW25E/graph.json","events_json":"https://pith.science/api/pith-number/C6AL6HGQAENSCXXKUACOMFW25E/events.json","paper":"https://pith.science/paper/C6AL6HGQ"},"agent_actions":{"view_html":"https://pith.science/pith/C6AL6HGQAENSCXXKUACOMFW25E","download_json":"https://pith.science/pith/C6AL6HGQAENSCXXKUACOMFW25E.json","view_paper":"https://pith.science/paper/C6AL6HGQ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2504.14870&json=true","fetch_graph":"https://pith.science/api/pith-number/C6AL6HGQAENSCXXKUACOMFW25E/graph.json","fetch_events":"https://pith.science/api/pith-number/C6AL6HGQAENSCXXKUACOMFW25E/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/C6AL6HGQAENSCXXKUACOMFW25E/action/timestamp_anchor","attest_storage":"https://pith.science/pith/C6AL6HGQAENSCXXKUACOMFW25E/action/storage_attestation","attest_author":"https://pith.science/pith/C6AL6HGQAENSCXXKUACOMFW25E/action/author_attestation","sign_citation":"https://pith.science/pith/C6AL6HGQAENSCXXKUACOMFW25E/action/citation_signature","submit_replication":"https://pith.science/pith/C6AL6HGQAENSCXXKUACOMFW25E/action/replication_record"}},"created_at":"2026-07-05T11:13:24.325223+00:00","updated_at":"2026-07-05T11:13:24.325223+00:00"}