{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:UJUDHYFTEFMULT2H5Q5TZP5ZIQ","short_pith_number":"pith:UJUDHYFT","schema_version":"1.0","canonical_sha256":"a26833e0b3215945cf47ec3b3cbfb94428e7cd1704bd2e5feb4d808470ee715f","source":{"kind":"arxiv","id":"2601.12538","version":1},"attestation_state":"computed","paper":{"title":"Agentic Reasoning for Large Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Agentic reasoning turns large language models into autonomous agents that plan, act, and adapt through interaction.","cross_cats":["cs.CL"],"primary_cat":"cs.AI","authors_text":"Cheng Qian, Chi Wang, Dongqi Fu, Duo Zhou, Gaotang Li, Hanghang Tong, Heng Ji, Hui Liu, Jiaru Zou, Jiaxuan You, Jingrui He, Liri Fang, Mengting Ai, Ruizhong Qiu, Tianxin Wei, Ting-Wei Li, Wenxuan Bao, Xianfeng Tang, Xiangru Tang, Xiao Lin, Xuying Ning, Yin Xiao, Yuji Zhang, Yunzhe Li, Yu Wang, Ze Yang, Zhichen Zeng, Zhining Liu, Zihao Li","submitted_at":"2026-01-18T18:58:23Z","abstract_excerpt":"Reasoning is a fundamental cognitive process underlying inference, problem-solving, and decision-making. While large language models (LLMs) demonstrate strong reasoning capabilities in closed-world settings, they struggle in open-ended and dynamic environments. Agentic reasoning marks a paradigm shift by reframing LLMs as autonomous agents that plan, act, and learn through continual interaction. In this survey, we organize agentic reasoning along three complementary dimensions. First, we characterize environmental dynamics through three layers: foundational agentic reasoning, which establishes"},"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":true,"formal_links_present":true},"canonical_record":{"source":{"id":"2601.12538","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-01-18T18:58:23Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"967c5045eca00d3ac7fe25a93633c3cd313726b3682652b641ae7f17ef71b777","abstract_canon_sha256":"5013063d844cf6041cedf97e55bd43dfe3fb2ae2622cedb34cfb4c122915e95a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:38:13.729267Z","signature_b64":"1Nz8Wyhxg0y3KjID0LkchPMAIJhZY5Hv9f5kyrOOf3xYX5HX8iO3rGswfUtgDj4H4Zr8u8rHSeyHQCFyZO/PAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a26833e0b3215945cf47ec3b3cbfb94428e7cd1704bd2e5feb4d808470ee715f","last_reissued_at":"2026-05-17T23:38:13.728637Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:38:13.728637Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Agentic Reasoning for Large Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Agentic reasoning turns large language models into autonomous agents that plan, act, and adapt through interaction.","cross_cats":["cs.CL"],"primary_cat":"cs.AI","authors_text":"Cheng Qian, Chi Wang, Dongqi Fu, Duo Zhou, Gaotang Li, Hanghang Tong, Heng Ji, Hui Liu, Jiaru Zou, Jiaxuan You, Jingrui He, Liri Fang, Mengting Ai, Ruizhong Qiu, Tianxin Wei, Ting-Wei Li, Wenxuan Bao, Xianfeng Tang, Xiangru Tang, Xiao Lin, Xuying Ning, Yin Xiao, Yuji Zhang, Yunzhe Li, Yu Wang, Ze Yang, Zhichen Zeng, Zhining Liu, Zihao Li","submitted_at":"2026-01-18T18:58:23Z","abstract_excerpt":"Reasoning is a fundamental cognitive process underlying inference, problem-solving, and decision-making. While large language models (LLMs) demonstrate strong reasoning capabilities in closed-world settings, they struggle in open-ended and dynamic environments. Agentic reasoning marks a paradigm shift by reframing LLMs as autonomous agents that plan, act, and learn through continual interaction. In this survey, we organize agentic reasoning along three complementary dimensions. First, we characterize environmental dynamics through three layers: foundational agentic reasoning, which establishes"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"This survey synthesizes agentic reasoning methods into a unified roadmap bridging thought and action, and outlines open challenges and future directions, including personalization, long-horizon interaction, world modeling, scalable multi-agent training, and governance for real-world deployment.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The assumption that the three complementary dimensions—foundational agentic reasoning, self-evolving agentic reasoning, and collective multi-agent reasoning—provide a comprehensive and non-overlapping organization of the entire field of agentic reasoning for LLMs.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"The survey structures agentic reasoning for LLMs into foundational, self-evolving, and collective multi-agent layers while distinguishing in-context orchestration from post-training optimization and reviewing applications across domains.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Agentic reasoning turns large language models into autonomous agents that plan, act, and adapt through interaction.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"fc6d4b9a2eb8132f0fa3230377ab97a1466d3c5f694c3bfd26184f62c72ad14a"},"source":{"id":"2601.12538","kind":"arxiv","version":1},"verdict":{"id":"6dc741c3-fe81-478b-b282-75cf171590ac","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-17T15:08:51.511537Z","strongest_claim":"This survey synthesizes agentic reasoning methods into a unified roadmap bridging thought and action, and outlines open challenges and future directions, including personalization, long-horizon interaction, world modeling, scalable multi-agent training, and governance for real-world deployment.","one_line_summary":"The survey structures agentic reasoning for LLMs into foundational, self-evolving, and collective multi-agent layers while distinguishing in-context orchestration from post-training optimization and reviewing applications across domains.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The assumption that the three complementary dimensions—foundational agentic reasoning, self-evolving agentic reasoning, and collective multi-agent reasoning—provide a comprehensive and non-overlapping organization of the entire field of agentic reasoning for LLMs.","pith_extraction_headline":"Agentic reasoning turns large language models into autonomous agents that plan, act, and adapt through interaction."},"references":{"count":300,"sample":[{"doi":"","year":2022,"title":"Chain-of-thought prompting elicits reasoning in large language models.Advances in neural information processing systems, 35:24824–24837","work_id":"3256b5a6-76d9-460c-ad77-5d232058ad6d","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2022,"title":"Least-to-Most Prompting Enables Complex Reasoning in Large Language Models","work_id":"7e58c111-4666-4996-b5ad-1c8efd433083","ref_index":2,"cited_arxiv_id":"2205.10625","is_internal_anchor":true},{"doi":"","year":2023,"title":"Pal: Program-aided language models","work_id":"3e5da501-d2ca-4bec-804c-85eef4857e6f","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2023,"title":"Tree of thoughts: Deliberate problem solving with large language models.Advances in neural information processing systems, 36:11809–11822","work_id":"7f3f800e-7253-4585-a6f5-f29d7ea04fe7","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2023,"title":"React: Synergizing reasoning and acting in language models","work_id":"db3d7fd5-c023-4dce-b5d9-acb8403ec05f","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":300,"snapshot_sha256":"e0a4d28892947179760dec61064bb5866627b7989a6070c9d7a98b8177111001","internal_anchors":56},"formal_canon":{"evidence_count":2,"snapshot_sha256":"3072f8d829590b5e948a5455fcf8da0c614ec01e8216aec9c48666ba075e1666"},"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":"2601.12538","created_at":"2026-05-17T23:38:13.728734+00:00"},{"alias_kind":"arxiv_version","alias_value":"2601.12538v1","created_at":"2026-05-17T23:38:13.728734+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2601.12538","created_at":"2026-05-17T23:38:13.728734+00:00"},{"alias_kind":"pith_short_12","alias_value":"UJUDHYFTEFMU","created_at":"2026-05-18T12:33:37.589309+00:00"},{"alias_kind":"pith_short_16","alias_value":"UJUDHYFTEFMULT2H","created_at":"2026-05-18T12:33:37.589309+00:00"},{"alias_kind":"pith_short_8","alias_value":"UJUDHYFT","created_at":"2026-05-18T12:33:37.589309+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":20,"internal_anchor_count":20,"sample":[{"citing_arxiv_id":"2605.14133","citing_title":"ClawForge: Generating Executable Interactive Benchmarks for Command-Line Agents","ref_index":100,"is_internal_anchor":true},{"citing_arxiv_id":"2604.20858","citing_title":"Mixture of Sequence: Theme-Aware Mixture-of-Experts for Long-Sequence Recommendation","ref_index":117,"is_internal_anchor":true},{"citing_arxiv_id":"2605.14133","citing_title":"ClawForge: Generating Executable Interactive Benchmarks for Command-Line Agents","ref_index":100,"is_internal_anchor":true},{"citing_arxiv_id":"2605.00200","citing_title":"Confidence Estimation in Automatic Short Answer Grading with LLMs","ref_index":34,"is_internal_anchor":true},{"citing_arxiv_id":"2605.13037","citing_title":"MAP: A Map-then-Act Paradigm for Long-Horizon Interactive Agent Reasoning","ref_index":33,"is_internal_anchor":true},{"citing_arxiv_id":"2604.03512","citing_title":"ActionNex: A Virtual Outage Manager for Cloud Computing","ref_index":14,"is_internal_anchor":true},{"citing_arxiv_id":"2605.06130","citing_title":"Skill1: Unified Evolution of Skill-Augmented Agents via Reinforcement Learning","ref_index":53,"is_internal_anchor":true},{"citing_arxiv_id":"2605.12004","citing_title":"Learning Agentic Policy from Action Guidance","ref_index":63,"is_internal_anchor":true},{"citing_arxiv_id":"2604.26615","citing_title":"TDD Governance for Multi-Agent Code Generation via Prompt Engineering","ref_index":23,"is_internal_anchor":true},{"citing_arxiv_id":"2605.09879","citing_title":"M2A: Synergizing Mathematical and Agentic Reasoning in Large Language Models","ref_index":39,"is_internal_anchor":true},{"citing_arxiv_id":"2605.10698","citing_title":"The Bystander Effect in Multi-Agent Reasoning: Quantifying Cognitive Loafing in Collaborative Interactions","ref_index":32,"is_internal_anchor":true},{"citing_arxiv_id":"2605.06130","citing_title":"Skill1: Unified Evolution of Skill-Augmented Agents via Reinforcement Learning","ref_index":53,"is_internal_anchor":true},{"citing_arxiv_id":"2605.00200","citing_title":"Confidence Estimation in Automatic Short Answer Grading with LLMs","ref_index":34,"is_internal_anchor":true},{"citing_arxiv_id":"2605.06130","citing_title":"Skill1: Unified Evolution of Skill-Augmented Agents via Reinforcement Learning","ref_index":53,"is_internal_anchor":true},{"citing_arxiv_id":"2604.05719","citing_title":"Hackers or Hallucinators? A Comprehensive Analysis of LLM-Based Automated Penetration Testing","ref_index":114,"is_internal_anchor":true},{"citing_arxiv_id":"2604.16646","citing_title":"Agentic Frameworks for Reasoning Tasks: An Empirical Study","ref_index":6,"is_internal_anchor":true},{"citing_arxiv_id":"2604.17821","citing_title":"WebUncertainty: Dual-Level Uncertainty Driven Planning and Reasoning For Autonomous Web Agent","ref_index":32,"is_internal_anchor":true},{"citing_arxiv_id":"2604.21027","citing_title":"HypEHR: Hyperbolic Modeling of Electronic Health Records for Efficient Question Answering","ref_index":78,"is_internal_anchor":true},{"citing_arxiv_id":"2605.02801","citing_title":"Reinforcement Learning for LLM-based Multi-Agent Systems through Orchestration Traces","ref_index":65,"is_internal_anchor":true},{"citing_arxiv_id":"2604.27351","citing_title":"Heterogeneous Scientific Foundation Model Collaboration","ref_index":4,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":2,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/UJUDHYFTEFMULT2H5Q5TZP5ZIQ","json":"https://pith.science/pith/UJUDHYFTEFMULT2H5Q5TZP5ZIQ.json","graph_json":"https://pith.science/api/pith-number/UJUDHYFTEFMULT2H5Q5TZP5ZIQ/graph.json","events_json":"https://pith.science/api/pith-number/UJUDHYFTEFMULT2H5Q5TZP5ZIQ/events.json","paper":"https://pith.science/paper/UJUDHYFT"},"agent_actions":{"view_html":"https://pith.science/pith/UJUDHYFTEFMULT2H5Q5TZP5ZIQ","download_json":"https://pith.science/pith/UJUDHYFTEFMULT2H5Q5TZP5ZIQ.json","view_paper":"https://pith.science/paper/UJUDHYFT","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2601.12538&json=true","fetch_graph":"https://pith.science/api/pith-number/UJUDHYFTEFMULT2H5Q5TZP5ZIQ/graph.json","fetch_events":"https://pith.science/api/pith-number/UJUDHYFTEFMULT2H5Q5TZP5ZIQ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/UJUDHYFTEFMULT2H5Q5TZP5ZIQ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/UJUDHYFTEFMULT2H5Q5TZP5ZIQ/action/storage_attestation","attest_author":"https://pith.science/pith/UJUDHYFTEFMULT2H5Q5TZP5ZIQ/action/author_attestation","sign_citation":"https://pith.science/pith/UJUDHYFTEFMULT2H5Q5TZP5ZIQ/action/citation_signature","submit_replication":"https://pith.science/pith/UJUDHYFTEFMULT2H5Q5TZP5ZIQ/action/replication_record"}},"created_at":"2026-05-17T23:38:13.728734+00:00","updated_at":"2026-05-17T23:38:13.728734+00:00"}