mlsanityData to 14 Oct 2026

The Danger of Overthinking: Examining the Reasoning-Action Dilemma in Agentic Tasks

arXiv 2502.08235 ↗First posted 12 Feb 2025cs.AIcited by 152

On the map: LLM reasoning → Test-time compute and efficient thinking

also in 1 more
Abstract (arXiv)

Large Reasoning Models (LRMs) represent a breakthrough in AI problem-solving capabilities, but their effectiveness in interactive environments can be limited. This paper introduces and analyzes overthinking in LRMs. A phenomenon where models favor extended internal reasoning chains over environmental interaction. Through experiments on software engineering tasks using SWE Bench Verified, we observe three recurring patterns: Analysis Paralysis, Rogue Actions, and Premature Disengagement. We propose a framework to study these behaviors, which correlates with human expert assessments, and analyze 4018 trajectories. We observe that higher overthinking scores correlate with decreased performance, with reasoning models exhibiting stronger tendencies toward overthinking compared to non-reasoning models. Our analysis reveals that simple efforts to mitigate overthinking in agentic environments, such as selecting the solution with the lower overthinking score, can improve model performance by almost 30% while reducing computational costs by 43%. These results suggest that mitigating overthinking has strong practical implications. We suggest that by leveraging native function-calling capabilities and selective reinforcement learning overthinking tendencies could be mitigated. We also open-source our evaluation framework and dataset to facilitate research in this direction at https://github.com/AlexCuadron/Overthinking.

Key facts. Cited by 152 papers in the map (data through 14 Oct 2026). Placed in Test-time compute and efficient thinking, one of 17 directions in LLM reasoning. Builds on 26 of its 53 references that are in the map. First posted on arXiv 12 Feb 2025.

Builds on

26 of 53 references.

Closest ancestors · 4

Same direction, newest first

  1. 2024Do NOT Think That Much for 2+3=? On the Overthinking of o1-Like LLMs30 Dec
  2. 2024Reasoning Paths Optimization: Learning to Reason and Explore From Diverse Paths7 Oct
  3. 2023Tree of Thoughts: Deliberate Problem Solving with Large Language Models17 May
  4. 2023Self-Refine: Iterative Refinement with Self-Feedback30 Mar

Other references · 22

  1. 2022Chain-of-Thought Prompting Elicits Reasoning in Large Language Modelscited by 12,459
  2. 2023Judging LLM-as-a-Judge with MT-Bench and Chatbot Arenacited by 8,626
  3. 2020Measuring Massive Multitask Language Understandingcited by 7,963
  4. 2025DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learningcited by 6,330
  5. 2022Large Language Models are Zero-Shot Reasonerscited by 4,374
  6. 2023SWE-bench: Can Language Models Resolve Real-World GitHub Issues?cited by 3,373
  7. 2023GPQA: A Graduate-Level Google-Proof Q&A Benchmarkcited by 2,879
  8. 2024Qwen2 Technical Reportcited by 2,120
  9. 2024SWE-agent: Agent-Computer Interfaces Enable Automated Software Engineeringcited by 1,737
  10. 2024OpenHands: An Open Platform for AI Software Developers as Generalist Agentscited by 1,002
  11. 2022A Survey on In-context Learningcited by 721
  12. 2024Executable Code Actions Elicit Better LLM Agentscited by 564
10 more
  1. 2025Search-o1: Agentic Search-Enhanced Large Reasoning Modelscited by 433
  2. 2024Agentless: Demystifying LLM-based Software Engineering Agentscited by 421
  3. 2025rStar-Math: Small LLMs Can Master Math Reasoning with Self-Evolved Deep Thinkingcited by 313
  4. 2025Towards Large Reasoning Models: A Survey of Reinforced Reasoning with Large Language Modelscited by 293
  5. 2024AI Agents That Mattercited by 219
  6. 2024Evaluation of OpenAI o1: Opportunities and Challenges of AGIcited by 127
  7. 2024CodeR: Issue Resolving with Multi-Agent and Task Graphscited by 91
  8. 2024MarsCode Agent: AI-native Automated Bug Fixingcited by 71
  9. 2024HyperAgent: Generalist Software Engineering Agents to Solve Coding Tasks at Scalecited by 61
  10. 2024Agentic Information Retrievalcited by 21

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 108; 2026: 44.

152 papers.

Sort:

By area:

  1. 2025Stop Overthinking: A Survey on Efficient Reasoning for Large Language Modelscited by 464
  2. 2025Towards Reasoning Era: A Survey of Long Chain-of-Thought for Reasoning Large Language Modelscited by 384
  3. 2025Dynamic Early Exit in Reasoning Modelscited by 219
  4. 2025A Survey of Efficient Reasoning for Large Reasoning Models: Language, Multimodality, and Beyondcited by 132
  5. 2025Process Reward Models That Thinkcited by 118
  6. 2025Thinkless: LLM Learns When to Thinkcited by 113
  7. 2025Efficient Reasoning Models: A Surveycited by 95
  8. 2025S-GRPO: Early Exit via Reinforcement Learning in Reasoning Modelscited by 87
  9. 2025ShorterBetter: Guiding Reasoning Models to Find Optimal Inference Length for Efficient Reasoningcited by 72
  10. 2025Wait, We Don't Need to "Wait"! Removing Thinking Tokens Improves Reasoning Efficiencycited by 66
  11. 2025SWE-Exp: Experience-Driven Software Issue Resolutioncited by 65
  12. 2025Safety in Large Reasoning Models: A Surveycited by 64
38 more
  1. 2025Efficient Inference for Large Reasoning Models: A Surveycited by 61
  2. 2025Missing Premise exacerbates Overthinking: Are Reasoning Models losing Critical Thinking Skill?cited by 58
  3. 2025Inverse Scaling in Test-Time Computecited by 50
  4. 2025Crosslingual Reasoning through Test-Time Scalingcited by 46
  5. 2025Think or Not Think: A Study of Explicit Thinking in Rule-Based Visual Reinforcement Fine-Tuningcited by 44
  6. 2025SciReplicate-Bench: Benchmarking LLMs in Agent-driven Algorithmic Reproduction from Research Paperscited by 43
  7. 2025Optimizing Length Compression in Large Reasoning Modelscited by 38
  8. 2025Making Slow Thinking Faster: Compressing LLM Chain-of-Thought via Step Entropycited by 38
  9. 2025ParaThinker: Native Parallel Thinking as a New Paradigm to Scale LLM Test-time Computecited by 38
  10. 2025ThinkSwitcher: When to Think Hard, When to Think Fastcited by 37
  11. 2025Don't Overthink It: A Survey of Efficient R1-style Large Reasoning Modelscited by 37
  12. 2025Harnessing the Reasoning Economy: A Survey of Efficient Reasoning for Large Language Modelscited by 36
  13. 2025Cost-of-Pass: An Economic Framework for Evaluating Language Modelscited by 35
  14. 2025OptimalThinkingBench: Evaluating Over and Underthinking in LLMscited by 35
  15. 2025Retro-Search: Exploring Untaken Paths for Deeper and Efficient Reasoningcited by 34
  16. 2025Mitigating Overthinking in Large Reasoning Models via Manifold Steeringcited by 33
  17. 2025100 Days After DeepSeek-R1: A Survey on Replication Studies and More Directions for Reasoning Language Modelscited by 30
  18. 2025Topology of Reasoning: Understanding Large Reasoning Models through Reasoning Graph Propertiescited by 29
  19. 2025Synergizing RAG and Reasoning: A Systematic Reviewcited by 27
  20. 2025Stable Reinforcement Learning for Efficient Reasoningcited by 27
  21. 2025Wait, Wait, Wait... Why Do Reasoning Models Loop?cited by 26
  22. 2025A Comprehensive Survey on Trustworthiness in Reasoning with Large Language Modelscited by 25
  23. 2025MathAgent: Leveraging a Mixture-of-Math-Agent Framework for Real-World Multimodal Mathematical Error Detectioncited by 24
  24. 2025Beyond Prompt Engineering: Robust Behavior Control in LLMs via Steering Target Atomscited by 24
  25. 2025How to Train Your Advisor: Steering Black-Box LLMs with Advisor Modelscited by 24
  26. 2025Position: Agent Should Invoke External Tools ONLY When Epistemically Necessarycited by 22
  27. 2025Language Mixing in Reasoning Language Models: Patterns, Impact, and Internal Causescited by 21
  28. 2025What Makes a Good Reasoning Chain? Uncovering Structural Patterns in Long Chain-of-Thought Reasoningcited by 21
  29. 2025The Lighthouse of Language: Enhancing LLM Agents via Critique-Guided Improvementcited by 20
  30. 2025THINK-Bench: Evaluating Thinking Efficiency and Chain-of-Thought Quality of Large Reasoning Modelscited by 19
  31. 2025Reasoning Models Hallucinate More: Factuality-Aware Reinforcement Learning for Large Reasoning Modelscited by 19
  32. 2025Do Language Models Mirror Human Confidence? Exploring Psychological Insights to Address Overconfidence in LLMscited by 19
  33. 2025Towards Concise and Adaptive Thinking in Large Reasoning Models: A Surveycited by 18
  34. 2025WebCoT: Enhancing Web Agent Reasoning by Reconstructing Chain-of-Thought in Reflection, Branching, and Rollbackcited by 17
  35. 2025OThink-R1: Intrinsic Fast/Slow Thinking Mode Switching for Over-Reasoning Mitigationcited by 17
  36. 2025Causal Sufficiency and Necessity Improves Chain-of-Thought Reasoningcited by 17
  37. 2025Reasoning on a Spectrum: Aligning LLMs to System 1 and System 2 Thinkingcited by 16
  38. 2025Two Experts Are All You Need for Steering Thinking: Reinforcing Cognitive Effort in MoE Reasoning Models Without Additional Trainingcited by 16

50 most cited of 152 All 152 (JSON)

Similar, not cited

Close in wording, no citation either way

  1. 2025Large Reasoning Models are not thinking straight: on the unreliability of thinking trajectories
  2. 2026Mitigating Overthinking in Large Reasoning Models via Difficulty-aware Reinforcement Learning
  3. 2025Think How to Think: Mitigating Overthinking with Autonomous Difficulty Cognition in Large Reasoning Models
  4. 2026Don't Overthink, Don't Underthink: Toward Adaptive Reasoning in Agentic AI
  5. 2026Mitigating Overthinking in Large Reasoning Language Models via Reasoning Path Deviation Monitoring
  6. 2025Let LRMs Break Free from Overthinking via Self-Braking Tuning
  7. 2025Rethinking the Illusion of Thinking
  8. 2026Constructing Pareto Frontiers of Optimal Reasoning Termination Methods for Overthinking in Reasoning Models