mlsanityData to 7 Oct 2026

Harnessing the Reasoning Economy: A Survey of Efficient Reasoning for Large Language Models

arXiv 2503.24377 ↗First posted 31 Mar 2025cs.CL cs.AIcited by 35

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

Abstract (arXiv)

Recent advancements in Large Language Models (LLMs) have significantly enhanced their ability to perform complex reasoning tasks, transitioning from fast and intuitive thinking (System 1) to slow and deep reasoning (System 2). While System 2 reasoning improves task accuracy, it often incurs substantial computational costs due to its slow thinking nature and inefficient or unnecessary reasoning behaviors. In contrast, System 1 reasoning is computationally efficient but leads to suboptimal performance. Consequently, it is critical to balance the trade-off between performance (benefits) and computational costs (budgets), giving rise to the concept of reasoning economy. In this survey, we provide a comprehensive analysis of reasoning economy in both the post-training and test-time inference stages of LLMs, encompassing i) the cause of reasoning inefficiency, ii) behavior analysis of different reasoning patterns, and iii) potential solutions to achieve reasoning economy. By offering actionable insights and highlighting open challenges, we aim to shed light on strategies for improving the reasoning economy of LLMs, thereby serving as a valuable resource for advancing research in this evolving area. We also provide a public repository to continually track developments in this fast-evolving field.

Key facts. Cited by 35 papers in the map (data through 7 Oct 2026). Placed in Test-time compute and efficient thinking, one of 17 directions in LLM reasoning. Builds on 122 of its 243 references that are in the map. First posted on arXiv 31 Mar 2025.

Builds on

122 of 243 references.

Closest ancestors · 36

Same direction, newest first

  1. 2025A Survey of Efficient Reasoning for Large Reasoning Models: Language, Multimodality, and Beyond27 Mar
  2. 2025Stop Overthinking: A Survey on Efficient Reasoning for Large Language Models20 Mar
  3. 2025L1: Controlling How Long A Reasoning Model Thinks With Reinforcement Learning6 Mar
  4. 2025Sampling-Efficient Test-Time Scaling: Self-Estimating the Best-of-N Sampling in Early Decoding3 Mar
  5. 2025How Well do LLMs Compress Their Own Chain-of-Thought? A Token Complexity Approach3 Mar
  6. 2025Thinking Slow, Fast: Scaling Inference Compute with Distilled Reasoners27 Feb
  7. 2025Efficient Test-Time Scaling via Self-Calibration25 Feb
  8. 2025Towards Thinking-Optimal Scaling of Test-Time Compute for LLM Reasoning25 Feb
  9. 2025The relationship between reasoning and performance in large language models--o3 (mini) thinks harder, not longer21 Feb
  10. 2025LightThinker: Thinking Step-by-Step Compression21 Feb
  11. 2025Inference-Time Computations for LLM Reasoning and Planning: A Benchmark and Insights18 Feb
  12. 2025Scaling Test-Time Compute Without Verification or RL is Suboptimal17 Feb
24 more
  1. 2025TokenSkip: Controllable Chain-of-Thought Compression in LLMs17 Feb
  2. 2025The Danger of Overthinking: Examining the Reasoning-Action Dilemma in Agentic Tasks12 Feb
  3. 2025Evolving LLMs' Self-Refinement Capability via Synergistic Training-Inference Optimization8 Feb
  4. 2025Scaling up Test-Time Compute with Latent Reasoning: A Recurrent Depth Approach7 Feb
  5. 2025Demystifying Long Chain-of-Thought Reasoning in LLMs5 Feb
  6. 2025s1: Simple test-time scaling31 Jan
  7. 2025Reward-Guided Speculative Decoding for Efficient LLM Reasoning31 Jan
  8. 2025Thoughts Are All Over the Place: On the Underthinking of o1-Like LLMs30 Jan
  9. 2024Do NOT Think That Much for 2+3=? On the Overthinking of o1-Like LLMs30 Dec
  10. 2024Token-Budget-Aware LLM Reasoning24 Dec
  11. 2024Compressed Chain of Thought: Efficient Reasoning Through Dense Representations17 Dec
  12. 2024TreeBoN: Enhancing Inference-Time Alignment with Speculative Tree-Search and Best-of-N Sampling18 Oct
  13. 2024O1 Replication Journey: A Strategic Progress Report -- Part 18 Oct
  14. 2024Adaptive Inference-Time Compute: LLMs Can Predict if They Can Do Better, Even Mid-Generation3 Oct
  15. 2024Training Language Models to Self-Correct via Reinforcement Learning19 Sep
  16. 2024Reasoning Aware Self-Consistency: Leveraging Reasoning Paths for Efficient LLM Sampling30 Aug
  17. 2024Path-Consistency with Prefix Enhancement for Efficient Inference in LLMs25 Aug
  18. 2024Make Every Penny Count: Difficulty-Adaptive Self-Consistency for Cost-Efficient Reasoning24 Aug
  19. 2024Scaling LLM Test-Time Compute Optimally can be More Effective than Scaling Model Parameters6 Aug
  20. 2024Concise Thoughts: Impact of Output Length on LLM Reasoning and Cost29 Jul
  21. 2024Escape Sky-high Cost: Early-stopping Self-Consistency for Multi-step Reasoning19 Jan
  22. 2024ReFT: Reasoning with Reinforced Fine-Tuning17 Jan
  23. 2023Tree of Thoughts: Deliberate Problem Solving with Large Language Models17 May
  24. 2023Self-Evaluation Guided Beam Search for Reasoning1 May

Other references · 86

  1. 2023GPT-4 Technical Reportcited by 18,132
  2. 2021Training Verifiers to Solve Math Word Problemscited by 9,743
  3. 2025DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learningcited by 6,146
  4. 2022Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedbackcited by 3,634
  5. 2023Let's Verify Step by Stepcited by 3,598
  6. 2025Search-R1: Training LLMs to Reason and Leverage Search Engines with Reinforcement Learningcited by 1,371
  7. 2025Kimi k1.5: Scaling Reinforcement Learning with LLMscited by 1,193
  8. 2023LIMA: Less Is More for Alignmentcited by 1,080
  9. 2024Qwen2.5-Math Technical Report: Toward Mathematical Expert Model via Self-Improvementcited by 1,057
  10. 2024SimPO: Simple Preference Optimization with a Reference-Free Rewardcited by 951
  11. 2024Large Language Monkeys: Scaling Inference Compute with Repeated Samplingcited by 946
  12. 2023Math-Shepherd: Verify and Reinforce LLMs Step-by-step without Human Annotationscited by 895
38 more
  1. 2023DSPy: Compiling Declarative Language Model Calls into Self-Improving Pipelinescited by 803
  2. 2024Training Large Language Models to Reason in a Continuous Latent Spacecited by 763
  3. 2023AGIEval: A Human-Centric Benchmark for Evaluating Foundation Modelscited by 701
  4. 2025SFT Memorizes, RL Generalizes: A Comparative Study of Foundation Model Post-trainingcited by 655
  5. 2025Vision-R1: Incentivizing Reasoning Capability in Multimodal Large Language Modelscited by 653
  6. 2023LLM-Pruner: On the Structural Pruning of Large Language Modelscited by 643
  7. 2024RouteLLM: Learning to Route LLMs with Preference Datacited by 633
  8. 2023Language Agent Tree Search Unifies Reasoning Acting and Planning in Language Modelscited by 551
  9. 2024Scaling and evaluating sparse autoencoderscited by 530
  10. 2025LIMO: Less is More for Reasoningcited by 488
  11. 2023Reinforced Self-Training (ReST) for Language Modelingcited by 486
  12. 2025Humanity's Last Examcited by 467
  13. 2025Video-R1: Reinforcing Video Reasoning in MLLMscited by 464
  14. 2024Improve Mathematical Reasoning in Language Models by Automated Process Supervisioncited by 426
  15. 2025The Lessons of Developing Process Reward Models in Mathematical Reasoningcited by 402
  16. 2025Towards Reasoning Era: A Survey of Long Chain-of-Thought for Reasoning Large Language Modelscited by 381
  17. 2024ShortGPT: Layers in Large Language Models are More Redundant Than You Expectcited by 369
  18. 2024Hybrid LLM: Cost-Efficient and Quality-Aware Query Routingcited by 354
  19. 2025From System 1 to System 2: A Survey of Reasoning Large Language Modelscited by 319
  20. 2024INSIDE: LLMs' Internal States Retain the Power of Hallucination Detectioncited by 317
  21. 2024Alignment faking in large language modelscited by 316
  22. 2024MoE-LLaVA: Mixture of Experts for Large Vision-Language Modelscited by 298
  23. 2025R1-VL: Learning to Reason with Multimodal Large Language Models via Step-wise Group Relative Policy Optimizationcited by 296
  24. 2025MM-Eureka: Exploring the Frontiers of Multimodal Reasoning with Rule-based Reinforcement Learningcited by 294
  25. 2025LMM-R1: Empowering 3B LMMs with Strong Reasoning Abilities Through Two-Stage Rule-Based RLcited by 244
  26. 2024DeepSeek-Prover: Advancing Theorem Proving in LLMs through Large-Scale Synthetic Datacited by 225
  27. 2024From Explicit CoT to Implicit CoT: Learning to Internalize CoT Step by Stepcited by 225
  28. 2024A Survey on Efficient Inference for Large Language Modelscited by 197
  29. 2022Defining and Characterizing Reward Hackingcited by 184
  30. 2025When More is Less: Understanding Chain-of-Thought Length in LLMscited by 182
  31. 2024Mulberry: Empowering MLLM with o1-like Reasoning and Reflection via Collective Monte Carlo Tree Searchcited by 171
  32. 2023Implicit Chain of Thought Reasoning via Knowledge Distillationcited by 165
  33. 2024Distilling System 2 into System 1cited by 152
  34. 2024Boosting Multimodal Large Language Models with Visual Tokens Withdrawal for Rapid Inferencecited by 137
  35. 2024Exploring the Reasoning Abilities of Multimodal Large Language Models (MLLMs): A Comprehensive Survey on Emerging Trends in Multimodal Reasoningcited by 136
  36. 2025LIMR: Less is More for RL Scalingcited by 136
  37. 2025MME-CoT: Benchmarking Chain-of-Thought in Large Multimodal Models for Reasoning Quality, Robustness, and Efficiencycited by 123
  38. 2025Optimizing Test-Time Compute via Meta Reinforcement Fine-Tuningcited by 121

50 most cited of 86 All 86 (JSON)

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 28; 2026: 7.

35 papers.

Sort:

By area:

  1. 2025Dynamic Early Exit in Reasoning Modelscited by 217
  2. 2025Acting Less is Reasoning More! Teaching Model to Act Efficientlycited by 80
  3. 2025Learn to Reason Efficiently with Adaptive Length-based Reward Shapingcited by 74
  4. 2025AdaCtrl: Towards Adaptive and Controllable Reasoning via Difficulty-Aware Budgetingcited by 50
  5. 2025Token Reduction Should Go Beyond Efficiency in Generative Models -- From Vision, Language to Multimodalitycited by 49
  6. 2025ToolOrchestra: Elevating Intelligence via Efficient Model and Tool Orchestrationcited by 44
  7. 2025Implicit Reasoning in Large Language Models: A Comprehensive Surveycited by 42
  8. 2025CompassVerifier: A Unified and Robust Verifier for LLMs Evaluation and Outcome Rewardcited by 39
  9. 2025Don't Overthink It: A Survey of Efficient R1-style Large Reasoning Modelscited by 36
  10. 2025Reasoning on a Budget: A Survey of Adaptive and Controllable Test-Time Compute in LLMscited by 31
  11. 2025Multimodal Spatial Reasoning in the Large Model Era: A Survey and Benchmarkscited by 31
  12. 2025R1-Compress: Long Chain-of-Thought Compression via Chunk Compression and Searchcited by 29
23 more
  1. 2025Let LRMs Break Free from Overthinking via Self-Braking Tuningcited by 27
  2. 2025Thinking Fast and Right: Balancing Accuracy and Reasoning Length with Adaptive Rewardscited by 20
  3. 2025The Evolution of Thought: Tracking LLM Overthinking via Reasoning Dynamics Analysiscited by 18
  4. 2026Learning to Self-Verify Makes Language Models Better Reasonerscited by 18
  5. 2025Two Experts Are All You Need for Steering Thinking: Reinforcing Cognitive Effort in MoE Reasoning Models Without Additional Trainingcited by 16
  6. 2025Don't Think Longer, Think Wisely: Optimizing Thinking Dynamics for Large Reasoning Modelscited by 11
  7. 2025Efficient Reasoning Through Suppression of Self-Affirmation Reflections in Large Reasoning Modelscited by 10
  8. 2025Revisiting Test-Time Scaling: A Survey and a Diversity-Aware Method for Efficient Reasoningcited by 8
  9. 2025Exploring and Exploiting the Inherent Efficiency within Large Reasoning Models for Self-Guided Efficiency Enhancementcited by 8
  10. 2026The Vision Wormhole: Latent-Space Communication in Heterogeneous Multi-Agent Systemscited by 8
  11. 2025ReliableMath: Benchmark of Reliable Mathematical Reasoning on Large Language Modelscited by 7
  12. 2025CoRE: Enhancing Metacognition with Label-free Self-evaluation in LRMscited by 5
  13. 2025SSR: Speculative Parallel Scaling Reasoning in Test-timecited by 4
  14. 2025Reasoning Can Hurt the Inductive Abilities of Large Language Modelscited by 4
  15. 2025Word Salad Chopper: Reasoning Models Waste A Ton Of Decoding Budget On Useless Repetitions, Self-Knowinglycited by 3
  16. 2025From Token to Action: State Machine Reasoning to Mitigate Overthinking in Information Retrievalcited by 2
  17. 2025EffiReason-Bench: A Unified Benchmark for Evaluating and Advancing Efficient Reasoning in Large Language Modelscited by 1
  18. 2025From Perception to Reasoning: Deep Thinking Empowers Multimodal Large Language Modelscited by 1
  19. 2026Segment-Level Attribution for Selective Learning of Long Reasoning Tracescited by 1
  20. 2026Dynamic Rollout Editing for Reducing Overthinking in RL-Trained Reasoning Modelscited by 1
  21. 2026</think> Doesn't Stop Reasoning: Analysis of Spurious CoT Terminationcited by 1
  22. 2026Can We Trust LLMs on Memristors? Diving into Reasoning Ability under Non-Idealitycited by 0
  23. 2026Test-time reasoning effort and unauthorized tool use in language-model agents: a prespecified equivalence studycited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2022Towards Reasoning in Large Language Models: A Survey
  2. 2025Advancing Reasoning in Large Language Models: Promising Methods and Approaches
  3. 2024Beyond Accuracy: Evaluating the Reasoning Behavior of Large Language Models -- A Survey
  4. 2024Inference Scaling vs Reasoning: An Empirical Analysis of Compute-Optimal LLM Problem-Solving
  5. 2025Efficient Inference for Large Reasoning Models: A Survey
  6. 2026The Reasoning Tax: Token Economics of LLM Reasoning Across Task Types and Deployment Contexts
  7. 2025From Efficiency to Adaptivity: A Deeper Look at Adaptive Reasoning in Large Language Models
  8. 2025Training Language Models to Reason Efficiently