mlsanityData to 14 Oct 2026

DeepAgent: A General Reasoning Agent with Scalable Toolsets

arXiv 2510.21618 ↗First posted 24 Oct 2025cs.AI cs.CL cs.IRcited by 71

On the map: LLM reasoning → Reasoning for agents, planning and code

also in 1 more
Abstract (arXiv)

Large reasoning models have demonstrated strong problem-solving abilities, yet real-world tasks often require external tools and long-horizon interactions. Existing agent frameworks typically follow predefined workflows, which limit autonomous and global task completion. In this paper, we introduce DeepAgent, an end-to-end deep reasoning agent that performs autonomous thinking, tool discovery, and action execution within a single, coherent reasoning process. To manage long-horizon interactions, we introduce an autonomous memory folding mechanism that compresses past interactions into structured episodic, working, and tool memories, reducing error accumulation while preserving critical information. To teach general-purpose tool use efficiently and stably, we develop an end-to-end reinforcement learning strategy, namely ToolPO, that leverages LLM-simulated APIs and applies tool-call advantage attribution to assign fine-grained credit to the tool invocation tokens. Extensive experiments on eight benchmarks, including general tool-use tasks (ToolBench, API-Bank, TMDB, Spotify, ToolHop) and downstream applications (ALFWorld, WebShop, GAIA, HLE), demonstrate that DeepAgent consistently outperforms baselines across both labeled-tool and open-set tool retrieval scenarios. The code and demo are available at https://github.com/RUC-NLPIR/DeepAgent.

Key facts. Cited by 71 papers in the map (data through 14 Oct 2026). Placed in Reasoning for agents, planning and code, one of 17 directions in LLM reasoning. Builds on 36 of its 80 references that are in the map. First posted on arXiv 24 Oct 2025.

Builds on

36 of 80 references.

Closest ancestors · 5

Same direction, newest first

  1. 2025Toward Effective Tool-Integrated Reasoning via Self-Evolved Preference Learning27 Sep
  2. 2025WebSailor-V2: Bridging the Chasm to Proprietary Agents via Synthetic Data and Scalable Reinforcement Learning16 Sep
  3. 2025VerlTool: Towards Holistic Agentic Reinforcement Learning with Tool Use1 Sep
  4. 2025ReTool: Reinforcement Learning for Strategic Tool Use in LLMs15 Apr
  5. 2024AgentGen: Enhancing Planning Abilities for Large Language Model based Agent via Environment and Task Generation1 Aug

Other references · 31

  1. 2022ReAct: Synergizing Reasoning and Acting in Language Modelscited by 8,695
  2. 2023Toolformer: Language Models Can Teach Themselves to Use Toolscited by 4,182
  3. 2023A Survey on Large Language Model based Autonomous Agentscited by 2,242
  4. 2024HybridFlow: A Flexible and Efficient RLHF Frameworkcited by 2,057
  5. 2022WebShop: Towards Scalable Real-World Web Interaction with Grounded Language Agentscited by 1,251
  6. 2020ALFWorld: Aligning Text and Embodied Environments for Interactive Learningcited by 1,205
  7. 2023GAIA: a benchmark for General AI Assistantscited by 1,128
  8. 2023API-Bank: A Comprehensive Benchmark for Tool-Augmented LLMscited by 591
  9. 2023C-Pack: Packed Resources For General Chinese Embeddingscited by 578
  10. 2024Do NOT Think That Much for 2+3=? On the Overthinking of o1-Like LLMscited by 570
  11. 2023AgentTuning: Enabling Generalized Agent Abilities for LLMscited by 338
  12. 2025Model Context Protocol (MCP): Landscape, Security Threats, and Future Research Directionscited by 265
19 more
  1. 2025WebDancer: Towards Autonomous Information Seeking Agencycited by 188
  2. 2024Imitate, Explore, and Self-Improve: A Reproduction Report on Slow-thinking Reasoning Systemscited by 164
  3. 2025Scaling Long-Horizon LLM Agent via Context-Foldingcited by 148
  4. 2025SimpleTIR: End-to-End Reinforcement Learning for Multi-Turn Tool-Integrated Reasoningcited by 146
  5. 2024O1 Replication Journey: A Strategic Progress Report -- Part 1cited by 143
  6. 2025ACON: Optimizing Context Compression for Long-horizon LLM Agentscited by 136
  7. 2025Towards Thinking-Optimal Scaling of Test-Time Compute for LLM Reasoningcited by 119
  8. 2024Diffusion Actor-Critic with Entropy Regulatorcited by 94
  9. 2024ToolGen: Unified Tool Retrieval and Calling via Generationcited by 74
  10. 2025AgentGym-RL: Training LLM Agents for Long-Horizon Decision Making through Multi-Turn Reinforcement Learningcited by 72
  11. 2025In-the-Flow Agentic System Optimization for Effective Planning and Tool Usecited by 72
  12. 2025Towards General Agentic Intelligence via Environment Scalingcited by 54
  13. 2025ToolHop: A Query-Driven Benchmark for Evaluating Large Language Models in Multi-Hop Tool Usecited by 50
  14. 2025Retrieval Models Aren't Tool-Savvy: Benchmarking Tool Retrieval for Large Language Modelscited by 47
  15. 2025Scaling Agents via Continual Pre-trainingcited by 40
  16. 2024RetroLLM: Empowering Large Language Models to Retrieve Fine-grained Evidence within Generationcited by 20
  17. 2025Hierarchical Document Refinement for Long-context Retrieval-augmented Generationcited by 16
  18. 2025FinSight: Towards Real-World Financial Deep Researchcited by 9
  19. 2025HierSearch: A Hierarchical Enterprise Deep Search Framework Integrating Local and Web Searchescited by 7

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 4; 2026: 67.

71 papers.

Sort:

By area:

  1. 2026Agent-World: Scaling Real-World Environment Synthesis for Evolving General Agent Intelligencecited by 54
  2. 2026A Survey of Agent Memory in the Second Half: Towards Self-Evolving and Long-Horizon Agentscited by 44
  3. 2026Stop Comparing LLM Agents Without Disclosing the Harnesscited by 34
  4. 2026ReCreate: Reasoning and Creating Domain Agents Driven by Experiencecited by 24
  5. 2026CoEvolve: Training LLM Agents via Agent-Data Mutual Evolutioncited by 23
  6. 2026OmniGAIA: Towards Native Omni-Modal AI Agentscited by 22
  7. 2026OS-Symphony: A Holistic Framework for Robust and Generalist Computer-Using Agentcited by 21
  8. 2026From Storage to Experience: A Survey on the Evolution of LLM Agent Memory Mechanismscited by 21
  9. 2026Self-Improvements in Modern Agentic Systems: A Surveycited by 19
  10. 2026ArenaRL: Scaling RL for Open-Ended Agents via Tournament-based Relative Rankingcited by 17
  11. 2026Marco DeepResearch: Unlocking Efficient Deep Research Agents via Verification-Centric Designcited by 16
  12. 2026Harness-R1: Learning to Edit Executable Runtime Harnesses from Agent Failure Trajectoriescited by 16
38 more
  1. 2026JIT-Agent: Scaling Harness Intelligence via Just-in-Time Harness Evolutioncited by 16
  2. 2026The Evolution of Tool Use in LLM Agents: From Single-Tool Call to Multi-Tool Orchestrationcited by 14
  3. 2026MemoBrain: Executive Memory as an Agentic Brain for Reasoningcited by 13
  4. 2025AI Meets Brain: Memory Systems from Cognitive Neuroscience to Autonomous Agentscited by 12
  5. 2026Mem2Evolve: Towards Self-Evolving Agents via Co-Evolutionary Capability Expansion and Experience Distillationcited by 12
  6. 2026Thinking with Map: Reinforced Parallel Map-Augmented Agent for Geolocalizationcited by 9
  7. 2026LaSER: Internalizing Explicit Reasoning into Latent Space for Dense Retrievalcited by 9
  8. 2026UMEM: Unified Memory Extraction and Management Framework for Generalizable Memorycited by 8
  9. 2026Unlocking Multimodal Document Intelligence: From Current Triumphs to Future Frontiers of Visual Document Retrievalcited by 8
  10. 2026SE-Search: Self-Evolving Search Agent via Memory and Dense Rewardcited by 8
  11. 2026TRACES: Proactive Safety Auditing for Multi-Turn LLM Agents via Trajectory-State Modelingcited by 8
  12. 2026Agentic Environment Engineering for Large Language Models: A Survey of Environment Modeling, Synthesis, Evaluation, and Applicationcited by 8
  13. 2025SciAgent: A Unified Multi-Agent System for Generalistic Scientific Reasoningcited by 7
  14. 2026CausalRM: Causal-Theoretic Reward Modeling for RLHF from Observational User Feedbackscited by 7
  15. 2026RoboAgent: Chaining Basic Capabilities for Embodied Task Planningcited by 7
  16. 2026ToolOmni: Enabling Open-World Tool Use via Agentic learning with Proactive Retrieval and Grounded Executioncited by 7
  17. 2025SAGE: Training Smart Any-Horizon Agents for Long Video Reasoning with Reinforcement Learningcited by 6
  18. 2026DLLM Agent: See Farther, Run Fastercited by 6
  19. 2026Yunjue Agent Tech Report: A Fully Reproducible, Zero-Start In-Situ Self-Evolving Agent System for Open-Ended Taskscited by 5
  20. 2026Table-as-Search: Formulate Long-Horizon Agentic Information Seeking as Table Completioncited by 5
  21. 2026Doctor-RAG: A Failure-Aware Repair Framework for Agentic Retrieval-Augmented Generationcited by 5
  22. 2026Bridging the Agent-World Gap: Text World Models for LLM-based Agentscited by 5
  23. 2026ET-Agent: Incentivizing Effective Tool-Integrated Reasoning Agent via Behavior Calibrationcited by 4
  24. 2026Can RL Improve Generalization of LLM Agents? An Empirical Studycited by 4
  25. 2026Seeing is Improving: Visual Feedback for Iterative Text Layout Refinementcited by 4
  26. 2026Agentic Tool Use in Large Language Modelscited by 4
  27. 2026APPO: Agentic Procedural Policy Optimizationcited by 4
  28. 2026From Atomic Actions to Standard Operating Procedures: Iterative Tool Optimization for Self-Evolving LLM Agentscited by 4
  29. 2026SkillForge: Evolving Verifiable Skills for Reinforcement Learning Agentscited by 4
  30. 2026Yunque DeepResearch Technical Reportcited by 3
  31. 2026LAMMI-Pathology: A Tool-Centric Bottom-Up LVLM-Agent Framework for Molecularly Informed Medical Intelligence in Pathologycited by 3
  32. 2026Mem-W: Latent Memory-Native GUI Agentscited by 3
  33. 2026HIPIF: Hierarchical Planning and Information Folding for Long-Horizon LLM Agent Learningcited by 3
  34. 2026Guided by Trajectories: Repairing and Rewarding Tool-Use Trajectories for Tool-Integrated Reasoningcited by 2
  35. 2026Dynamic Mixed-Precision Routing for Efficient Multi-step LLM Interactioncited by 2
  36. 2026ACE: Pluggable Adaptive Context Elasticizer across Agentscited by 2
  37. 2026Rubrics on Trial: Evolving Rubrics from a Single Query via Synthetic Pairwise Evidencecited by 2
  38. 2026HyMem: Hierarchical Context Management for Long-Horizon Agents via Information Isolationcited by 2

50 most cited of 71 All 71 (JSON)

Similar, not cited

Close in wording, no citation either way

  1. 2025AlphaApollo: A System for Deep Agentic Reasoning
  2. 2026Deep Reasoning in General Purpose Agents via Structured Meta-Cognition
  3. 2026ToolVerse: Unlocking Massive Environments and Long-Horizon Tasks for Agentic Reinforcement Learning
  4. 2025AutoTool: Dynamic Tool Selection and Integration for Agentic Reasoning
  5. 2025Agent0: Unleashing Self-Evolving Agents from Zero Data via Tool-Integrated Reasoning
  6. 2025On Generalization in Agentic Tool Calling: CoreThink Agentic Reasoner and MAVEN Dataset
  7. 2025Agent-as-Tool: A Study on the Hierarchical Decision Making with Reinforcement Learning
  8. 2026Evolving from Tool User to Creator via Training-Free Experience Reuse in Multimodal Reasoning