mlsanityData to 14 Oct 2026

Towards Concise and Adaptive Thinking in Large Reasoning Models: A Survey

arXiv 2507.09662 ↗First posted 13 Jul 2025cs.AI cs.CLcited by 18

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

Abstract (arXiv)

Large reasoning models (LRMs) like OpenAI o1 and DeepSeek R1 have demonstrated impressive performance on complex reasoning tasks like mathematics and programming with long Chain-of-Thought (CoT) reasoning sequences (slow-thinking), compared with traditional large language models (fast-thinking). However, these reasoning models also face a huge challenge that generating unnecessarily lengthy and redundant reasoning chains even for trivial questions. This phenomenon leads to a significant waste of inference resources, increases the response time for simple queries, and hinders the practical application of LRMs in real-world products. To this end, it is crucial to shorten lengthy reasoning chains and learn adaptive reasoning between fast and slow thinking based on input difficulty. In this survey, we provide a comprehensive overview of recent progress in concise and adaptive thinking for efficient reasoning of LRMs, including methodologies, benchmarks, and challenges for future exploration. We hope this survey can help researchers quickly understand the landscape of this field and inspire novel adaptive thinking ideas to facilitate better usage of LRMs.

Key facts. Cited by 18 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 190 of its 218 references that are in the map. First posted on arXiv 13 Jul 2025.

Builds on

190 of 218 references.

Closest ancestors · 93

Same direction, newest first

  1. 2025Think How to Think: Mitigating Overthinking with Autonomous Difficulty Cognition in Large Reasoning Models3 Jul
  2. 2025Reasoning on a Budget: A Survey of Adaptive and Controllable Test-Time Compute in LLMs2 Jul
  3. 2025Do Thinking Tokens Help or Trap? Towards More Efficient Large Reasoning Model30 Jun
  4. 2025Is Long-to-Short a Free Lunch? Investigating Inconsistency and Reasoning Efficiency in LRMs24 Jun
  5. 2025ConciseHint: Boosting Efficient Reasoning via Continuous Concise Hints during Generation23 Jun
  6. 2025Reasoning about Uncertainty: Do Reasoning Models Know When They Don't Know?22 Jun
  7. 2025Optimizing Length Compression in Large Reasoning Models17 Jun
  8. 2025Steering LLM Thinking with Budget Guidance16 Jun
  9. 2025MiniMax-M1: Scaling Test-Time Compute Efficiently with Lightning Attention16 Jun
  10. 2025QFFT, Question-Free Fine-Tuning for Adaptive Reasoning15 Jun
  11. 2025Efficient Reasoning Through Suppression of Self-Affirmation Reflections in Large Reasoning Models14 Jun
  12. 2025ReCUT: Balancing Reasoning Length and Accuracy in LLMs via Stepwise Trails and Preference Optimization12 Jun
38 more
  1. 2025Fast on the Easy, Deep on the Hard: Efficient Reasoning via Powered Length Penalty12 Jun
  2. 2025On Reasoning Strength Planning in Large Reasoning Models10 Jun
  3. 2025Wait, We Don't Need to "Wait"! Removing Thinking Tokens Improves Reasoning Efficiency10 Jun
  4. 2025Overclocking LLM Reasoning: Monitoring and Controlling Thinking Path Lengths in LLMs8 Jun
  5. 2025How Far Are We from Optimal Reasoning Efficiency?8 Jun
  6. 2025Just Enough Thinking: Efficient Reasoning with Adaptive Length Penalties Reinforcement Learning5 Jun
  7. 2025Does Thinking More always Help? Mirage of Test-Time Scaling in Reasoning Models4 Jun
  8. 2025OThink-R1: Intrinsic Fast/Slow Thinking Mode Switching for Over-Reasoning Mitigation3 Jun
  9. 2025AlphaOne: Reasoning Models Thinking Slow and Fast at Test Time30 May
  10. 2025Mitigating Overthinking in Large Reasoning Models via Manifold Steering28 May
  11. 2025THINK-Bench: Evaluating Thinking Efficiency and Chain-of-Thought Quality of Large Reasoning Models28 May
  12. 2025The Price of a Second Thought: On the Evaluation of Reasoning Efficiency in Large Language Models28 May
  13. 2025Don't Think Longer, Think Wisely: Optimizing Thinking Dynamics for Large Reasoning Models27 May
  14. 2025R2R: Efficiently Navigating Divergent Reasoning Paths with Small-Large Model Token Routing27 May
  15. 2025Thinker: Learning to Think Fast and Slow27 May
  16. 2025Self-Route: Automatic Mode Switching via Capability Estimation for Efficient Reasoning27 May
  17. 2025Adaptive Deep Reasoning: Triggering Deep Thinking When Needed26 May
  18. 2025Done Is Better than Perfect: Unlocking Efficient Reasoning by Structured Multi-Turn Decomposition26 May
  19. 2025Route to Reason: Adaptive Routing for LLM and Reasoning Strategy Selection26 May
  20. 2025The Overthinker's DIET: Cutting Token Calories with DIfficulty-AwarE Training25 May
  21. 2025AdaCtrl: Towards Adaptive and Controllable Reasoning via Difficulty-Aware Budgeting24 May
  22. 2025Think or Not? Exploring Thinking Efficiency in Large Reasoning Models via an Information-Theoretic Lens23 May
  23. 2025VeriThinker: Learning to Verify Makes Reasoning Model Efficient23 May
  24. 2025Don't Overthink it. Preferring Shorter Thinking Chains for Improved LLM Reasoning23 May
  25. 2025Think Silently, Think Fast: Dynamic Latent Compression of LLM Reasoning Chains22 May
  26. 2025The First Impression Problem: Internal Bias Triggers Overthinking in Reasoning Models22 May
  27. 2025Incentivizing Dual Process Thinking for Efficient Large Language Model Reasoning22 May
  28. 2025Plan and Budget: Effective and Efficient Test-Time Scaling on Reasoning Large Language Models22 May
  29. 2025Hunyuan-TurboS: Advancing Large Language Models through Mamba-Transformer Synergy and Adaptive Chain-of-Thought21 May
  30. 2025When to Continue Thinking: Adaptive Thinking Mode Switching for Efficient Reasoning21 May
  31. 2025When Can Large Reasoning Models Save Thinking? Mechanistic Analysis of Behavioral Divergence in Reasoning21 May
  32. 2025Prolonged Reasoning Is Not All You Need: Certainty-Based Adaptive Routing for Efficient LLM/MLLM Reasoning21 May
  33. 2025Think Only When You Need with Large Hybrid-Reasoning Models20 May
  34. 2025Let LRMs Break Free from Overthinking via Self-Braking Tuning20 May
  35. 2025ThinkSwitcher: When to Think Hard, When to Think Fast20 May
  36. 2025Optimizing Anytime Reasoning via Budget Relative Policy Optimization19 May
  37. 2025AdaptThink: Reasoning Models Can Learn When to Think19 May
  38. 2025Thinkless: LLM Learns When to Think19 May

50 newest of 93 All 93 (JSON)

Other references · 97

  1. 2024The Llama 3 Herd of Modelscited by 14,715
  2. 2021Training Verifiers to Solve Math Word Problemscited by 10,109
  3. 2021Evaluating Large Language Models Trained on Codecited by 8,320
  4. 2021Measuring Mathematical Problem Solving With the MATH Datasetcited by 5,371
  5. 2018Think you have Solved Question Answering? Try ARC, the AI2 Reasoning Challengecited by 4,886
  6. 2018HotpotQA: A Dataset for Diverse, Explainable Multi-hop Question Answeringcited by 3,754
  7. 2021Program Synthesis with Large Language Modelscited by 3,637
  8. 2023SWE-bench: Can Language Models Resolve Real-World GitHub Issues?cited by 3,373
  9. 2023GPQA: A Graduate-Level Google-Proof Q&A Benchmarkcited by 2,879
  10. 2018Can a Suit of Armor Conduct Electricity? A New Dataset for Open Book Question Answeringcited by 2,164
  11. 2024LiveCodeBench: Holistic and Contamination Free Evaluation of Large Language Models for Codecited by 2,004
  12. 2018CommonsenseQA: A Question Answering Challenge Targeting Commonsense Knowledgecited by 1,822
38 more
  1. 2022Challenging BIG-Bench Tasks and Whether Chain-of-Thought Can Solve Themcited by 1,807
  2. 2022Solving Quantitative Reasoning Problems with Language Modelscited by 1,392
  3. 2023Is Your Code Generated by ChatGPT Really Correct? Rigorous Evaluation of Large Language Models for Code Generationcited by 1,391
  4. 2025Kimi k1.5: Scaling Reinforcement Learning with LLMscited by 1,210
  5. 2024OlympiadBench: A Challenging Benchmark for Promoting AGI with Olympiad-Level Bilingual Multimodal Scientific Problemscited by 1,136
  6. 2023Steering Llama 2 via Contrastive Activation Additioncited by 1,085
  7. 2019DROP: A Reading Comprehension Benchmark Requiring Discrete Reasoning Over Paragraphscited by 903
  8. 2024Length-Controlled AlpacaEval: A Simple Way to Debias Automatic Evaluatorscited by 897
  9. 2021Are NLP Models really able to Solve Simple Math Word Problems?cited by 878
  10. 2021Did Aristotle Use a Laptop? A Question Answering Benchmark with Implicit Reasoning Strategiescited by 743
  11. 2021Cross-Task Generalization via Natural Language Crowdsourcing Instructionscited by 699
  12. 2023Language Models are Super Mario: Absorbing Abilities from Homologous Models as a Free Lunchcited by 694
  13. 2024RouteLLM: Learning to Route LLMs with Preference Datacited by 682
  14. 2019MathQA: Towards Interpretable Math Word Problem Solving with Operation-Based Formalismscited by 662
  15. 2024BigCodeBench: Benchmarking Code Generation with Diverse Function Calls and Complex Instructionscited by 546
  16. 2024From Crowdsourced Data to High-Quality Benchmarks: Arena-Hard and BenchBuilder Pipelinecited by 500
  17. 2024TravelPlanner: A Benchmark for Real-World Planning with Language Agentscited by 392
  18. 2021MiniF2F: a cross-system benchmark for formal Olympiad-level mathematicscited by 391
  19. 2024Measuring short-form factuality in large language modelscited by 373
  20. 2024Hybrid LLM: Cost-Efficient and Quality-Aware Query Routingcited by 372
  21. 2021A Diverse Corpus for Evaluating and Developing English Math Word Problem Solverscited by 318
  22. 2025OpenThoughts: Data Recipes for Reasoning Modelscited by 304
  23. 2025Chain of Draft: Thinking Faster by Writing Lesscited by 297
  24. 2025MathArena: Evaluating LLMs on Uncontaminated Math Competitionscited by 297
  25. 2022Teaching Small Language Models to Reasoncited by 284
  26. 2025DeepMath-103K: A Large-Scale, Challenging, Decontaminated, and Verifiable Mathematical Dataset for Advancing Reasoningcited by 280
  27. 2024LiveBench: A Challenging, Contamination-Limited LLM Benchmarkcited by 262
  28. 2025Training Language Models to Reason Efficientlycited by 247
  29. 2024From Explicit CoT to Implicit CoT: Learning to Internalize CoT Step by Stepcited by 235
  30. 2024WildTeaming at Scale: From In-the-Wild Jailbreaks to (Adversarially) Safer Language Modelscited by 223
  31. 2020Retrieve, Program, Repeat: Complex Knowledge Base Question Answering via Alternate Meta-learningcited by 216
  32. 2023ProofNet: Autoformalizing and Formally Proving Undergraduate-Level Mathematicscited by 192
  33. 2024Let's Think Dot by Dot: Hidden Computation in Transformer Language Modelscited by 191
  34. 2025When More is Less: Understanding Chain-of-Thought Length in LLMscited by 183
  35. 2025Light-R1: Curriculum SFT, DPO and RL for Long COT from Scratch and Beyondcited by 175
  36. 2025AIMO-2 Winning Solution: Building State-of-the-Art Mathematical Reasoning Models with OpenMathReasoning datasetcited by 174
  37. 2025Seed1.5-Thinking: Advancing Superb Reasoning Models with Reinforcement Learningcited by 161
  38. 2024Distilling System 2 into System 1cited by 159

50 most cited of 97 All 97 (JSON)

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 10; 2026: 8.

18 papers.

Sort:

By area:

  1. 2025ParaThinker: Native Parallel Thinking as a New Paradigm to Scale LLM Test-time Computecited by 38
  2. 2025Reinforcement Learning Meets Large Language Models: A Survey of Advancements and Applications Across the LLM Lifecyclecited by 35
  3. 2025BudgetThinker: Empowering Budget-aware LLM Reasoning with Control Tokenscited by 29
  4. 2025Trace Length is a Simple Uncertainty Signal in Reasoning Modelscited by 19
  5. 2025ARES: Multimodal Adaptive Reasoning via Difficulty-Aware Token-Level Entropy Shapingcited by 17
  6. 2026Self-Compression of Chain-of-Thought via Multi-Agent Reinforcement Learningcited by 13
  7. 2025From Efficiency to Adaptivity: A Deeper Look at Adaptive Reasoning in Large Language Modelscited by 12
  8. 2025The Sequential Edge: Inverse-Entropy Voting Beats Parallel Self-Consistency at Matched Computecited by 11
  9. 2026Are Latent Reasoning Models Easily Interpretable?cited by 9
  10. 2026Think Dense, Not Long: Dynamic Decoupled Conditional Advantage for Efficient Reasoningcited by 5
  11. 2025ARM2: Adaptive Reasoning Model with Vision Understanding and Executable Codecited by 4
  12. 2025Concise Reasoning in the Lens of Lagrangian Optimizationcited by 3
6 more
  1. 2026Not All Errors Are Equal: Consequence-Aware Reasoning Compute Allocationcited by 3
  2. 2026QA-Merging: Query-Adaptive Reasoning via Layer Selective Model Mergingcited by 1
  3. 2026InfoDensity: Rewarding Information-Dense Traces for Efficient Reasoningcited by 1
  4. 2025MMD-Thinker: Adaptive Multi-Dimensional Thinking for Multimodal Misinformation Detectioncited by 0
  5. 2026Policy Split: Incentivizing Dual-Mode Exploration in LLM Reinforcement with Dual-Mode Entropy Regularizationcited by 0
  6. 2026When Compression Helps and When It Hurts: Condition-Aware Analysis of Chain-of-Thought Distillationcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025A Survey of Efficient Reasoning for Large Reasoning Models: Language, Multimodality, and Beyond
  2. 2025Towards Reasoning Era: A Survey of Long Chain-of-Thought for Reasoning Large Language Models
  3. 2025Don't Overthink It: A Survey of Efficient R1-style Large Reasoning Models
  4. 2025Marco-o1 v2: Towards Widening The Distillation Bottleneck for Reasoning Models
  5. 2025Exploring the System 1 Thinking Capability of Large Reasoning Models
  6. 2025Efficient Reasoning Models: A Survey
  7. 2025One Missing Piece for Open-Source Reasoning Models: A Dataset to Mitigate Cold-Starting Short CoT LLMs in RL
  8. 2025From System 1 to System 2: A Survey of Reasoning Large Language Models