mlsanityData to 7 Oct 2026

Do NOT Think That Much for 2+3=? On the Overthinking of o1-Like LLMs

arXiv 2412.21187 ↗First posted 30 Dec 2024cs.CLcited by 564

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

Abstract (arXiv)

The remarkable performance of models like the OpenAI o1 can be attributed to their ability to emulate human-like long-time thinking during inference. These models employ extended chain-of-thought (CoT) processes, exploring multiple strategies to enhance problem-solving capabilities. However, a critical question remains: How to intelligently and efficiently scale computational resources during testing. This paper presents the first comprehensive study on the prevalent issue of overthinking in these models, where excessive computational resources are allocated for simple problems with minimal benefit. We introduce novel efficiency metrics from both outcome and process perspectives to evaluate the rational use of computational resources by o1-like models. Using a self-training paradigm, we propose strategies to mitigate overthinking, streamlining reasoning processes without compromising accuracy. Experimental results show that our approach successfully reduces computational overhead while preserving model performance across a range of testsets with varying difficulty levels, such as GSM8K, MATH500, GPQA, and AIME.

Key facts. Cited by 564 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 42 of its 52 references that are in the map. First posted on arXiv 30 Dec 2024.

Builds on

42 of 52 references.

Closest ancestors · 12

Same direction, newest first

  1. 2024Token-Budget-Aware LLM Reasoning24 Dec
  2. 2024Learning How Hard to Think: Input-Adaptive Allocation of LM Computation7 Oct
  3. 2024Adaptive Inference-Time Compute: LLMs Can Predict if They Can Do Better, Even Mid-Generation3 Oct
  4. 2024Training Language Models to Self-Correct via Reinforcement Learning19 Sep
  5. 2024Make Every Penny Count: Difficulty-Adaptive Self-Consistency for Cost-Efficient Reasoning24 Aug
  6. 2024Toward Self-Improvement of LLMs via Imagination, Searching, and Criticizing18 Apr
  7. 2024CriticBench: Benchmarking LLMs for Critique-Correct Reasoning22 Feb
  8. 2024Escape Sky-high Cost: Early-stopping Self-Consistency for Multi-step Reasoning19 Jan
  9. 2023Think before you speak: Training Language Models With Pause Tokens3 Oct
  10. 2023Alphazero-like Tree-Search can Guide Large Language Model Decoding and Training29 Sep
  11. 2023Tree of Thoughts: Deliberate Problem Solving with Large Language Models17 May
  12. 2022STaR: Bootstrapping Reasoning With Reasoning28 Mar

Other references · 30

  1. 2022Chain-of-Thought Prompting Elicits Reasoning in Large Language Modelscited by 12,302
  2. 2021Training Verifiers to Solve Math Word Problemscited by 9,743
  3. 2023Direct Preference Optimization: Your Language Model is Secretly a Reward Modelcited by 7,916
  4. 2022Self-Consistency Improves Chain of Thought Reasoning in Language Modelscited by 5,601
  5. 2021Measuring Mathematical Problem Solving With the MATH Datasetcited by 5,203
  6. 2022Large Language Models are Zero-Shot Reasonerscited by 4,323
  7. 2021Finetuned Language Models Are Zero-Shot Learnerscited by 3,747
  8. 2023Let's Verify Step by Stepcited by 3,598
  9. 2023GPQA: A Graduate-Level Google-Proof Q&A Benchmarkcited by 2,748
  10. 2023Improving Factuality and Reasoning in Language Models through Multiagent Debatecited by 1,739
  11. 2023Encouraging Divergent Thinking in Large Language Models through Multi-Agent Debatecited by 995
  12. 2024SimPO: Simple Preference Optimization with a Reference-Free Rewardcited by 951
18 more
  1. 2023Graph of Thoughts: Solving Elaborate Problems with Large Language Modelscited by 873
  2. 2024Training Large Language Models to Reason in a Continuous Latent Spacecited by 763
  3. 2023Plan-and-Solve Prompting: Improving Zero-Shot Chain-of-Thought Reasoning by Large Language Modelscited by 550
  4. 2024Improve Mathematical Reasoning in Language Models by Automated Process Supervisioncited by 426
  5. 2022Large Language Models Are Reasoning Teacherscited by 318
  6. 2021A Diverse Corpus for Evaluating and Developing English Math Word Problem Solverscited by 314
  7. 2024Iterative Reasoning Preference Optimizationcited by 244
  8. 2022Making Large Language Models Better Reasoners with Step-Aware Verifiercited by 231
  9. 2024When Can LLMs Actually Correct Their Own Mistakes? A Critical Survey of Self-Correction of LLMscited by 230
  10. 2024Provably Mitigating Overoptimization in RLHF: Your SFT Loss is Implicitly an Adversarial Regularizercited by 109
  11. 2024Common 7B Language Models Already Possess Strong Math Capabilitiescited by 108
  12. 2023Examining Inter-Consistency of Large Language Models Collaboration: An In-depth Analysis via Debatecited by 94
  13. 2024Prover-Verifier Games improve legibility of LLM outputscited by 79
  14. 2024Self-Generated Critiques Boost Reward Modeling for Language Modelscited by 69
  15. 2024Better Instruction-Following Through Minimum Bayes Riskcited by 19
  16. 2024Automatic Curriculum Expert Iteration for Reliable LLM Reasoningcited by 16
  17. 2024Improving Minimum Bayes Risk Decoding with Multi-Promptcited by 13
  18. 2024Assessing the Creativity of LLMs in Proposing Novel Solutions to Mathematical Problemscited by 9

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 390; 2026: 174.

564 papers.

Sort:

By area:

  1. 2025Understanding R1-Zero-Like Training: A Critical Perspectivecited by 1,276
  2. 2025Kimi k1.5: Scaling Reinforcement Learning with LLMscited by 1,193
  3. 2025SimpleRL-Zoo: Investigating and Taming Zero Reinforcement Learning for Open Base Models in the Wildcited by 505
  4. 2025Stop Overthinking: A Survey on Efficient Reasoning for Large Language Modelscited by 456
  5. 2025Search-o1: Agentic Search-Enhanced Large Reasoning Modelscited by 423
  6. 2025L1: Controlling How Long A Reasoning Model Thinks With Reinforcement Learningcited by 381
  7. 2025Towards Reasoning Era: A Survey of Long Chain-of-Thought for Reasoning Large Language Modelscited by 381
  8. 2025From System 1 to System 2: A Survey of Reasoning Large Language Modelscited by 319
  9. 2025The Illusion of Thinking: Understanding the Strengths and Limitations of Reasoning Models via the Lens of Problem Complexitycited by 298
  10. 2025Chain of Draft: Thinking Faster by Writing Lesscited by 290
  11. 2025OpenThoughts: Data Recipes for Reasoning Modelscited by 287
  12. 2025O1-Pruner: Length-Harmonizing Fine-Tuning for O1-Like Reasoning Pruningcited by 278
38 more
  1. 2025DeepMath-103K: A Large-Scale, Challenging, Decontaminated, and Verifiable Mathematical Dataset for Advancing Reasoningcited by 261
  2. 2025Training Language Models to Reason Efficientlycited by 242
  3. 2025Dynamic Early Exit in Reasoning Modelscited by 217
  4. 2025A Survey on Test-Time Scaling in Large Language Models: What, How, Where, and How Well?cited by 215
  5. 2025Deep Think with Confidencecited by 199
  6. 2025ThinkPrune: Pruning Long Chain-of-Thought of LLMs via Reinforcement Learningcited by 187
  7. 2025When More is Less: Understanding Chain-of-Thought Length in LLMscited by 182
  8. 2025A Survey of Reinforcement Learning for Large Reasoning Modelscited by 175
  9. 2025CoT-Valve: Length-Compressible Chain-of-Thought Tuningcited by 167
  10. 2025Thoughts Are All Over the Place: On the Underthinking of o1-Like LLMscited by 162
  11. 2025DAST: Difficulty-Adaptive Slow-Thinking for Large Reasoning Modelscited by 154
  12. 2025Reasoning Models Know When They're Right: Probing Hidden States for Self-Verificationcited by 154
  13. 2025The Danger of Overthinking: Examining the Reasoning-Action Dilemma in Agentic Taskscited by 150
  14. 2025AdaptThink: Reasoning Models Can Learn When to Thinkcited by 141
  15. 2025A Survey of Efficient Reasoning for Large Reasoning Models: Language, Multimodality, and Beyondcited by 131
  16. 2025LLM Post-Training: A Deep Dive into Reasoning Large Language Modelscited by 129
  17. 2025Right Question is Already Half the Answer: Fully Unsupervised LLM Reasoning Incentivizationcited by 124
  18. 2025Optimizing Test-Time Compute via Meta Reinforcement Fine-Tuningcited by 121
  19. 2025Towards Thinking-Optimal Scaling of Test-Time Compute for LLM Reasoningcited by 118
  20. 2025Process Reward Models That Thinkcited by 118
  21. 2025Between Underthinking and Overthinking: An Empirical Study of Reasoning Length and correctness in LLMscited by 115
  22. 2025Towards System 2 Reasoning in LLMs: Learning How to Think With Meta Chain-of-Thoughtcited by 108
  23. 2025A Survey of Frontiers in LLM Reasoning: Inference Scaling, Learning to Reason, and Agentic Systemscited by 108
  24. 2025Thinkless: LLM Learns When to Thinkcited by 107
  25. 2025MedReason: Eliciting Factual Medical Reasoning Steps in LLMs via Knowledge Graphscited by 105
  26. 2025Self-Training Elicits Concise Reasoning in Large Language Modelscited by 97
  27. 2025OverThink: Slowdown Attacks on Reasoning LLMscited by 94
  28. 2025Efficient Reasoning Models: A Surveycited by 93
  29. 2025VerlTool: Towards Holistic Agentic Reinforcement Learning with Tool Usecited by 92
  30. 2025GuardReasoner: Towards Reasoning-based LLM Safeguardscited by 89
  31. 2025S-GRPO: Early Exit via Reinforcement Learning in Reasoning Modelscited by 87
  32. 2025AdaCoT: Pareto-Optimal Adaptive Chain-of-Thought Triggering via Reinforcement Learningcited by 86
  33. 2025Scaling Test-Time Compute Without Verification or RL is Suboptimalcited by 82
  34. 2025An Empirical Study on Eliciting and Improving R1-like Reasoning Modelscited by 78
  35. 2025Efficient Test-Time Scaling via Self-Calibrationcited by 74
  36. 2025ShorterBetter: Guiding Reasoning Models to Find Optimal Inference Length for Efficient Reasoningcited by 70
  37. 2025Revisiting the Test-Time Scaling of o1-like Models: Do they Truly Possess Test-Time Scaling Capabilities?cited by 69
  38. 2025DeepAgent: A General Reasoning Agent with Scalable Toolsetscited by 69

50 most cited of 564 All 564 (JSON)

Similar, not cited

Close in wording, no citation either way

  1. 2024A Comparative Study on Reasoning Patterns of OpenAI's o1 Model
  2. 2024O1 Replication Journey -- Part 2: Surpassing O1-preview through Simple Distillation, Big Progress or Bitter Lesson?
  3. 2026Thinking Past the Answer: Evaluating Harmful Overthinking in Large Reasoning Models
  4. 2025The Virtues of Brevity: Avoid Overthinking in Parallel Test-Time Reasoning
  5. 2025Enhancing Reasoning Abilities of Small LLMs with Cognitive Alignment
  6. 2025Think Twice: Enhancing LLM Reasoning by Scaling Multi-round Test-time Thinking
  7. 2025Explore Briefly, Then Decide: Mitigating LLM Overthinking via Cumulative Entropy Regulation
  8. 2025Thinking Short and Right Over Thinking Long: Serving LLM Reasoning Efficiently and Accurately