mlsanityData to 7 Oct 2026

Token-Budget-Aware LLM Reasoning

arXiv 2412.18547 ↗First posted 24 Dec 2024cs.CL cs.AI cs.LGcited by 242

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

Abstract (arXiv)

Reasoning is critical for large language models (LLMs) to excel in a wide range of tasks. While methods like Chain-of-Thought (CoT) reasoning and enhance LLM performance by decomposing problems into intermediate steps, they also incur significant overhead in token usage, leading to increased costs. We find that the reasoning process of current LLMs is unnecessarily lengthy and it can be compressed by including a reasonable token budget in the prompt, but the choice of token budget plays a crucial role in the actual compression effectiveness. We then propose a token-budget-aware LLM reasoning framework that dynamically adjusts the number of reasoning tokens based on the reasoning complexity of each problem. Experiments show that our method effectively reduces token costs in CoT reasoning with only a slight performance reduction, offering a practical solution to balance efficiency and accuracy in LLM reasoning. Code: https://github.com/GeniusHTX/TALE

Key facts. Cited by 242 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 31 of its 45 references that are in the map. First posted on arXiv 24 Dec 2024.

Builds on

31 of 45 references.

Closest ancestors · 4

Same direction, newest first

  1. 2024Language Models are Hidden Reasoners: Unlocking Latent Reasoning Capabilities via Self-Rewarding6 Nov
  2. 2024Concise Thoughts: Impact of Output Length on LLM Reasoning and Cost29 Jul
  3. 2024Reasoning in Token Economies: Budget-Aware Evaluation of LLM Reasoning Strategies10 Jun
  4. 2023Alphazero-like Tree-Search can Guide Large Language Model Decoding and Training29 Sep

Other references · 27

  1. 2022Training language models to follow instructions with human feedbackcited by 15,712
  2. 2024The Llama 3 Herd of Modelscited by 14,192
  3. 2021LoRA: Low-Rank Adaptation of Large Language Modelscited by 14,084
  4. 2021Training Verifiers to Solve Math Word Problemscited by 9,743
  5. 2023Direct Preference Optimization: Your Language Model is Secretly a Reward Modelcited by 7,916
  6. 2022Self-Consistency Improves Chain of Thought Reasoning in Language Modelscited by 5,601
  7. 2021Program Synthesis with Large Language Modelscited by 3,509
  8. 2022Challenging BIG-Bench Tasks and Whether Chain-of-Thought Can Solve Themcited by 1,767
  9. 2022Fast Inference from Transformers via Speculative Decodingcited by 1,615
  10. 2022Least-to-Most Prompting Enables Complex Reasoning in Large Language Modelscited by 1,397
  11. 2023Reasoning with Language Model is Planning with World Modelcited by 846
  12. 2023Large Language Modelscited by 774
15 more
  1. 2024Training Large Language Models to Reason in a Continuous Latent Spacecited by 763
  2. 2019CodeSearchNet Challenge: Evaluating the State of Semantic Code Searchcited by 638
  3. 2024Hybrid LLM: Cost-Efficient and Quality-Aware Query Routingcited by 354
  4. 2022Large Language Models Are Reasoning Teacherscited by 318
  5. 2023Towards Revealing the Mystery behind Chain of Thought: A Theoretical Perspectivecited by 293
  6. 2023Faithful Chain-of-Thought Reasoningcited by 285
  7. 2021AI Chains: Transparent and Controllable Human-AI Interaction by Chaining Large Language Model Promptscited by 265
  8. 2024Monte Carlo Tree Search Boosts Reasoning via Iterative Preference Learningcited by 238
  9. 2022Making Large Language Models Better Reasoners with Step-Aware Verifiercited by 231
  10. 2024MathBench: Evaluating the Theory and Application Proficiency of LLMs with a Hierarchical Mathematics Benchmarkcited by 103
  11. 2023Response Length Perception and Sequence Scheduling: An LLM-Empowered LLM Inference Pipelinecited by 86
  12. 2024Visual Agents as Fast and Slow Thinkerscited by 53
  13. 2020PALM: Pre-training an Autoencoding&Autoregressive Language Model for Context-conditioned Generationcited by 42
  14. 2023What's the Magic Word? A Control Theory of LLM Promptingcited by 41
  15. 2022SeqZero: Few-shot Compositional Semantic Parsing with Sequential Prompts and Zero-shot Modelscited by 38

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2024: 2; 2025: 146; 2026: 94.

242 papers.

Sort:

By area:

  1. 2024Do NOT Think That Much for 2+3=? On the Overthinking of o1-Like LLMscited by 564
  2. 2025Stop Overthinking: A Survey on Efficient Reasoning for Large Language Modelscited by 456
  3. 2025Towards Reasoning Era: A Survey of Long Chain-of-Thought for Reasoning Large Language Modelscited by 381
  4. 2025From System 1 to System 2: A Survey of Reasoning Large Language Modelscited by 319
  5. 2025O1-Pruner: Length-Harmonizing Fine-Tuning for O1-Like Reasoning Pruningcited by 278
  6. 2025Dynamic Early Exit in Reasoning Modelscited by 217
  7. 2025When More is Less: Understanding Chain-of-Thought Length in LLMscited by 182
  8. 2025A Survey of Context Engineering for Large Language Modelscited by 120
  9. 2025Between Underthinking and Overthinking: An Empirical Study of Reasoning Length and correctness in LLMscited by 115
  10. 2025Thinkless: LLM Learns When to Thinkcited by 107
  11. 2025How Well do LLMs Compress Their Own Chain-of-Thought? A Token Complexity Approachcited by 95
  12. 2025OverThink: Slowdown Attacks on Reasoning LLMscited by 94
38 more
  1. 2025Efficient Reasoning Models: A Surveycited by 93
  2. 2025S-GRPO: Early Exit via Reinforcement Learning in Reasoning Modelscited by 87
  3. 2025AdaCoT: Pareto-Optimal Adaptive Chain-of-Thought Triggering via Reinforcement Learningcited by 86
  4. 2025ShorterBetter: Guiding Reasoning Models to Find Optimal Inference Length for Efficient Reasoningcited by 70
  5. 2025Quantization Hurts Reasoning? An Empirical Study on Quantized Reasoning Modelscited by 68
  6. 2025Effectively Controlling Reasoning Models through Thinking Interventioncited by 63
  7. 2025Learning When to Think: Shaping Adaptive Reasoning in R1-Style Models via Multi-Stage RLcited by 59
  8. 2025Efficient Inference for Large Reasoning Models: A Surveycited by 58
  9. 2024Efficiently Scaling LLM Reasoning with Certaindexcited by 56
  10. 2025InftyThink: Breaking the Length Limits of Long-Context Reasoning in Large Language Modelscited by 53
  11. 2025PolyMath: Evaluating Mathematical Reasoning in Multilingual Contextscited by 52
  12. 2025R-KV: Redundancy-aware KV Cache Compression for Reasoning Modelscited by 52
  13. 2025AdaCtrl: Towards Adaptive and Controllable Reasoning via Difficulty-Aware Budgetingcited by 50
  14. 2025Token Reduction Should Go Beyond Efficiency in Generative Models -- From Vision, Language to Multimodalitycited by 49
  15. 2025Inverse Scaling in Test-Time Computecited by 49
  16. 2025SelfBudgeter: Adaptive Token Allocation for Efficient LLM Reasoningcited by 47
  17. 2025Budget-Aware Tool Use Enables Effective Agent Scalingcited by 42
  18. 2025ARM: Adaptive Reasoning Modelcited by 38
  19. 2025Answer Convergence as a Signal for Early Stopping in Reasoningcited by 38
  20. 2025Optimizing Length Compression in Large Reasoning Modelscited by 37
  21. 2025Don't Overthink It: A Survey of Efficient R1-style Large Reasoning Modelscited by 36
  22. 2025Harnessing the Reasoning Economy: A Survey of Efficient Reasoning for Large Language Modelscited by 35
  23. 2025Prolonged Reasoning Is Not All You Need: Certainty-Based Adaptive Routing for Efficient LLM/MLLM Reasoningcited by 34
  24. 2025ConCISE: Confidence-guided Compression in Step-by-step Efficient Reasoningcited by 33
  25. 2025Think or Not? Exploring Thinking Efficiency in Large Reasoning Models via an Information-Theoretic Lenscited by 32
  26. 2025Efficient Agents: Building Effective Agents While Reducing Costcited by 32
  27. 2025Beyond GPT-5: Making LLMs Cheaper and Better via Performance-Efficiency Optimized Routingcited by 32
  28. 2025Plan and Budget: Effective and Efficient Test-Time Scaling on Reasoning Large Language Modelscited by 30
  29. 2025Steering LLM Thinking with Budget Guidancecited by 30
  30. 2025Lower Bounds for Chain-of-Thought Reasoning in Hard-Attention Transformerscited by 28
  31. 2025R2R: Efficiently Navigating Divergent Reasoning Paths with Small-Large Model Token Routingcited by 28
  32. 2025PixelThink: Towards Efficient Chain-of-Pixel Reasoningcited by 28
  33. 2026Single-Agent LLMs Outperform Multi-Agent Systems on Multi-Hop Reasoning Under Equal Thinking Token Budgetscited by 28
  34. 2025Optimizing Anytime Reasoning via Budget Relative Policy Optimizationcited by 27
  35. 2025Let LRMs Break Free from Overthinking via Self-Braking Tuningcited by 27
  36. 2025SPEED-RL: Faster Training of Reasoning Models via Online Curriculum Learningcited by 27
  37. 2025Promoting Efficient Reasoning with Verifiable Stepwise Rewardcited by 27
  38. 2025BudgetThinker: Empowering Budget-aware LLM Reasoning with Control Tokenscited by 27

50 most cited of 242 All 242 (JSON)

Similar, not cited

Close in wording, no citation either way

  1. 2025TokenSkip: Controllable Chain-of-Thought Compression in LLMs
  2. 2024Extending Token Computation for LLM Reasoning
  3. 2026Dual-Track CoT: Budget-Aware Stepwise Guidance for Small LMs
  4. 2025Reasoning on a Budget: A Survey of Adaptive and Controllable Test-Time Compute in LLMs
  5. 2026Avoiding Overthinking and Underthinking: Curriculum-Aware Budget Scheduling for LLMs
  6. 2025Revisiting LLM Reasoning via Information Bottleneck
  7. 2025Not All Tokens Matter: Towards Efficient LLM Reasoning via Token Significance in Reinforcement Learning
  8. 2026Zipping the Thought: When and How Compressed Reasoning Data Works in LLM Post-Training