{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:ENZAIGGZJ4JTKBLKCYEFQFD366","short_pith_number":"pith:ENZAIGGZ","schema_version":"1.0","canonical_sha256":"23720418d94f1335056a160858147bf7a9c34236b645bacc436093d577b8a443","source":{"kind":"arxiv","id":"2510.05592","version":2},"attestation_state":"computed","paper":{"title":"In-the-Flow Agentic System Optimization for Effective Planning and Tool Use","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL","cs.LG","cs.MA"],"primary_cat":"cs.AI","authors_text":"Haoxiang Zhang, James Zou, Jianwen Xie, Pan Lu, Seungju Han, Sheng Liu, Yejin Choi, Yu Zhang, Zhuofeng Li","submitted_at":"2025-10-07T05:32:44Z","abstract_excerpt":"Outcome-driven reinforcement learning has advanced reasoning in large language models (LLMs), but prevailing tool-augmented approaches train a single, monolithic policy that interleaves thoughts and tool calls under full context; this scales poorly with long horizons and diverse tools and generalizes weakly to new scenarios. Agentic systems offer a promising alternative by decomposing work across specialized modules, yet most remain training-free or rely on offline training decoupled from the live dynamics of multi-turn interaction. We introduce AgentFlow, a trainable, in-the-flow agentic fram"},"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":"2510.05592","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2025-10-07T05:32:44Z","cross_cats_sorted":["cs.CL","cs.LG","cs.MA"],"title_canon_sha256":"1cc56bf8f0f6c28689d899eed5955865f116316562a157e8796b09070b5ada26","abstract_canon_sha256":"82e9483b1af753c432322d61434add349d32bff2be4472819456e33e58d51cb3"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-23T01:23:29.429659Z","signature_b64":"ApVx4pqEFHuFOxZfygNRO7tCHvmYlBCLXHguhEYSefFTnqtQUlfVmnvvOBcwaXMAozHHHoWmPmLrAPeflncHDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"23720418d94f1335056a160858147bf7a9c34236b645bacc436093d577b8a443","last_reissued_at":"2026-07-23T01:23:29.428699Z","signature_status":"signed_v1","first_computed_at":"2026-07-23T01:23:29.428699Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"In-the-Flow Agentic System Optimization for Effective Planning and Tool Use","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL","cs.LG","cs.MA"],"primary_cat":"cs.AI","authors_text":"Haoxiang Zhang, James Zou, Jianwen Xie, Pan Lu, Seungju Han, Sheng Liu, Yejin Choi, Yu Zhang, Zhuofeng Li","submitted_at":"2025-10-07T05:32:44Z","abstract_excerpt":"Outcome-driven reinforcement learning has advanced reasoning in large language models (LLMs), but prevailing tool-augmented approaches train a single, monolithic policy that interleaves thoughts and tool calls under full context; this scales poorly with long horizons and diverse tools and generalizes weakly to new scenarios. Agentic systems offer a promising alternative by decomposing work across specialized modules, yet most remain training-free or rely on offline training decoupled from the live dynamics of multi-turn interaction. We introduce AgentFlow, a trainable, in-the-flow agentic fram"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2510.05592","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/2510.05592/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":"2510.05592","created_at":"2026-07-23T01:23:29.429144+00:00"},{"alias_kind":"arxiv_version","alias_value":"2510.05592v2","created_at":"2026-07-23T01:23:29.429144+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2510.05592","created_at":"2026-07-23T01:23:29.429144+00:00"},{"alias_kind":"pith_short_12","alias_value":"ENZAIGGZJ4JT","created_at":"2026-07-23T01:23:29.429144+00:00"},{"alias_kind":"pith_short_16","alias_value":"ENZAIGGZJ4JTKBLK","created_at":"2026-07-23T01:23:29.429144+00:00"},{"alias_kind":"pith_short_8","alias_value":"ENZAIGGZ","created_at":"2026-07-23T01:23:29.429144+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":11,"internal_anchor_count":11,"sample":[{"citing_arxiv_id":"2606.12087","citing_title":"FORT-Searcher: Synthesizing Shortcut-Resistant Search Tasks for Training Deep Search Agents","ref_index":61,"is_internal_anchor":true},{"citing_arxiv_id":"2606.10385","citing_title":"Beyond Absolute Imitation: Anchored Residual Guidance for Privileged On-Policy Distillation","ref_index":18,"is_internal_anchor":true},{"citing_arxiv_id":"2606.05009","citing_title":"DAR: Deontic Reasoning with Agentic Harnesses","ref_index":30,"is_internal_anchor":true},{"citing_arxiv_id":"2605.18772","citing_title":"Improving Retrieval-Augmented Generation without Taxonomy-based Error Categorization","ref_index":2,"is_internal_anchor":true},{"citing_arxiv_id":"2605.18396","citing_title":"NEWTON: Agentic Planning for Physically Grounded Video Generation","ref_index":19,"is_internal_anchor":true},{"citing_arxiv_id":"2603.16876","citing_title":"Multi-Modal Multi-Agent Reinforcement Learning for Radiology Report Generation","ref_index":63,"is_internal_anchor":true},{"citing_arxiv_id":"2605.08715","citing_title":"AgentForesight: Online Auditing for Early Failure Prediction in Multi-Agent Systems","ref_index":27,"is_internal_anchor":true},{"citing_arxiv_id":"2605.11706","citing_title":"GRAFT: Graph-Tokenized LLMs for Tool Planning","ref_index":21,"is_internal_anchor":true},{"citing_arxiv_id":"2605.08715","citing_title":"AgentForesight: Online Auditing for Early Failure Prediction in Multi-Agent Systems","ref_index":27,"is_internal_anchor":true},{"citing_arxiv_id":"2605.07462","citing_title":"The Moltbook Files: A Harmless Slopocalypse or Humanity's Last Experiment","ref_index":45,"is_internal_anchor":true},{"citing_arxiv_id":"2605.07339","citing_title":"Tools as Continuous Flow for Evolving Agentic Reasoning","ref_index":28,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/ENZAIGGZJ4JTKBLKCYEFQFD366","json":"https://pith.science/pith/ENZAIGGZJ4JTKBLKCYEFQFD366.json","graph_json":"https://pith.science/api/pith-number/ENZAIGGZJ4JTKBLKCYEFQFD366/graph.json","events_json":"https://pith.science/api/pith-number/ENZAIGGZJ4JTKBLKCYEFQFD366/events.json","paper":"https://pith.science/paper/ENZAIGGZ"},"agent_actions":{"view_html":"https://pith.science/pith/ENZAIGGZJ4JTKBLKCYEFQFD366","download_json":"https://pith.science/pith/ENZAIGGZJ4JTKBLKCYEFQFD366.json","view_paper":"https://pith.science/paper/ENZAIGGZ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2510.05592&json=true","fetch_graph":"https://pith.science/api/pith-number/ENZAIGGZJ4JTKBLKCYEFQFD366/graph.json","fetch_events":"https://pith.science/api/pith-number/ENZAIGGZJ4JTKBLKCYEFQFD366/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/ENZAIGGZJ4JTKBLKCYEFQFD366/action/timestamp_anchor","attest_storage":"https://pith.science/pith/ENZAIGGZJ4JTKBLKCYEFQFD366/action/storage_attestation","attest_author":"https://pith.science/pith/ENZAIGGZJ4JTKBLKCYEFQFD366/action/author_attestation","sign_citation":"https://pith.science/pith/ENZAIGGZJ4JTKBLKCYEFQFD366/action/citation_signature","submit_replication":"https://pith.science/pith/ENZAIGGZJ4JTKBLKCYEFQFD366/action/replication_record"}},"created_at":"2026-07-23T01:23:29.429144+00:00","updated_at":"2026-07-23T01:23:29.429144+00:00"}