mlsanityData to 7 Oct 2026

D-CORE: Incentivizing Task Decomposition in Large Reasoning Models for Complex Tool Use

arXiv 2602.02160 ↗First posted 2 Feb 2026cs.CLcited by 0

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

Abstract (arXiv)

Effective tool use and reasoning are essential capabilities for large reasoning models~(LRMs) to address complex real-world problems. Through empirical analysis, we identify that current LRMs lack the capability of sub-task decomposition in complex tool use scenarios, leading to Lazy Reasoning. To address this, we propose a two-stage training framework D-CORE~(underlineDecomposing tasks and underlineComposing underlineReasoning processes) that first incentivize the LRMs' task decomposition reasoning capability via self-distillation, followed by diversity-aware reinforcement learning~(RL) to restore LRMs' reflective reasoning capability. D-CORE achieves robust tool-use improvements across diverse benchmarks and model scales. Experiments on BFCLv3 demonstrate superiority of our method: D-CORE-8B reaches 77.7% accuracy, surpassing the best-performing 8B model by 5.7%. Meanwhile, D-CORE-14B establishes a new state-of-the-art at 79.3%, outperforming 70B models despite being 5× smaller. The source code is available at https://github.com/alibaba/EfficientAI.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in Reasoning for agents, planning and code, one of 17 directions in LLM reasoning. Builds on 37 of its 55 references that are in the map. First posted on arXiv 2 Feb 2026.

Builds on

37 of 55 references.

Closest ancestors · 2

Same direction, newest first

  1. 2025Tool-Star: Empowering LLM-Brained Multi-Tool Reasoner via Reinforcement Learning22 May
  2. 2025Nemotron-Research-Tool-N1: Exploring Tool-Using Language Models with Reinforced Reasoning25 Apr

Other references · 35

  1. 2024DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Modelscited by 8,851
  2. 2025Qwen3 Technical Reportcited by 8,195
  3. 2025DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learningcited by 6,146
  4. 2021Measuring Mathematical Problem Solving With the MATH Datasetcited by 5,203
  5. 2024HybridFlow: A Flexible and Efficient RLHF Frameworkcited by 1,980
  6. 2023WebArena: A Realistic Web Environment for Building Autonomous Agentscited by 1,908
  7. 2022Least-to-Most Prompting Enables Complex Reasoning in Large Language Modelscited by 1,397
  8. 2023Gorilla: Large Language Model Connected with Massive APIscited by 1,302
  9. 2024τ-bench: A Benchmark for Tool-Agent-User Interaction in Real-World Domainscited by 1,280
  10. 2025Does Reinforcement Learning Really Incentivize Reasoning Capacity in LLMs Beyond the Base Model?cited by 948
  11. 2022Large Language Models Can Self-Improvecited by 656
  12. 2025SFT Memorizes, RL Generalizes: A Comparative Study of Foundation Model Post-trainingcited by 655
23 more
  1. 2025BrowseComp: A Simple Yet Challenging Benchmark for Browsing Agentscited by 616
  2. 2025τ2-Bench: Evaluating Conversational Agents in a Dual-Control Environmentcited by 558
  3. 2022Decomposed Prompting: A Modular Approach for Solving Complex Taskscited by 554
  4. 2025Beyond the 80/20 Rule: High-Entropy Minority Tokens Drive Effective Reinforcement Learning for LLM Reasoningcited by 540
  5. 2025Cognitive Behaviors that Enable Self-Improving Reasoners, or, Four Habits of Highly Effective STaRscited by 373
  6. 2025ToolRL: Reward is All Tool Learning Needscited by 342
  7. 2025WebThinker: Empowering Large Reasoning Models with Deep Research Capabilitycited by 320
  8. 2025The Illusion of Thinking: Understanding the Strengths and Limitations of Reasoning Models via the Lens of Problem Complexitycited by 298
  9. 2025Reasoning with Exploration: An Entropy Perspectivecited by 277
  10. 2024ToolACE: Winning the Points of LLM Function Callingcited by 204
  11. 2025APIGen-MT: Agentic Pipeline for Multi-Turn Data Generation via Simulated Agent-Human Interplaycited by 158
  12. 2025Thought Anchors: Which LLM Reasoning Steps Matter?cited by 126
  13. 2023T-Eval: Evaluating the Tool Utilization Capability of Large Language Models Step by Stepcited by 91
  14. 2025Understanding Reasoning in Thinking Language Models via Steering Vectorscited by 69
  15. 2025ACEBench: Who Wins the Match Point in Tool Usage?cited by 66
  16. 2024Hammer: Robust Function-Calling for On-Device Language Models via Function Maskingcited by 53
  17. 2025ComplexFuncBench: Exploring Multi-Step and Constrained Function Calling under Long-Context Scenariocited by 53
  18. 2025START: Self-taught Reasoner with Toolscited by 47
  19. 2024Facilitating Multi-turn Function Calling for LLMs via Compositional Instruction Tuningcited by 40
  20. 2025Magnet: Multi-turn Tool-use Data Synthesis and Distillation via Graph Translationcited by 37
  21. 2025Not All Thoughts are Generated Equal: Efficient LLM Reasoning via Multi-Turn Reinforcement Learningcited by 16
  22. 2025Why Distillation can Outperform Zero-RL: The Role of Flexible Reasoningcited by 15
  23. 2024Mixture-of-Instructions: Aligning Large Language Models via Mixture Promptingcited by 4

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026TInR: Exploring Tool-Internalized Reasoning in Large Language Models
  2. 2025Step-wise Policy for Rare-tool Knowledge (SPaRK): Offline RL that Drives Diverse Tool Use in LLMs
  3. 2026Discovery and Reinforcement of Tool-Integrated Reasoning Chains via Rollout Trees
  4. 2025Toward Effective Tool-Integrated Reasoning via Self-Evolved Preference Learning
  5. 2026Teaching Thinking Models to Reason with Tools: A Full-Pipeline Recipe for Tool-Integrated Reasoning
  6. 2025AutoTIR: Autonomous Tools Integrated Reasoning via Reinforcement Learning
  7. 2026Evolving from Tool User to Creator via Training-Free Experience Reuse in Multimodal Reasoning
  8. 2025Divide-Then-Aggregate: An Efficient Tool Learning Method via Parallel Tool Invocation