mlsanityData to 14 Oct 2026

Large Reasoning Models are not thinking straight: on the unreliability of thinking trajectories

arXiv 2507.00711 ↗First posted 1 Jul 2025cs.LGcited by 10

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

Abstract (arXiv)

Large Language Models (LLMs) trained via Reinforcement Learning (RL) have recently achieved impressive results on reasoning benchmarks. Yet, growing evidence shows that these models often generate longer but ineffective chains of thought (CoTs), calling into question whether benchmark gains reflect real reasoning improvements. We present new evidence of overthinking, where models disregard correct solutions even when explicitly provided, instead continuing to generate unnecessary reasoning steps that often lead to incorrect conclusions. Experiments on three state-of-the-art models using the AIME2024 math benchmark reveal critical limitations in these models ability to integrate corrective information, posing new challenges for achieving robust and interpretable reasoning.

Key facts. Cited by 10 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 27 of its 38 references that are in the map. First posted on arXiv 1 Jul 2025.

Builds on

27 of 38 references.

Closest ancestors · 9

Same direction, newest first

  1. 2025Reasoning Models Can Be Effective Without Thinking14 Apr
  2. 2025Demystifying Long Chain-of-Thought Reasoning in LLMs5 Feb
  3. 2025s1: Simple test-time scaling31 Jan
  4. 2025Thoughts Are All Over the Place: On the Underthinking of o1-Like LLMs30 Jan
  5. 2025T1: Advancing Language Model Reasoning through Reinforcement Learning and Inference Scaling20 Jan
  6. 2025Monte Carlo Tree Search for Comprehensive Exploration in LLM-Based Automatic Heuristic Design15 Jan
  7. 2024Do NOT Think That Much for 2+3=? On the Overthinking of o1-Like LLMs30 Dec
  8. 2024Interpretable Contrastive Monte Carlo Tree Search Reasoning2 Oct
  9. 2024Toward Self-Improvement of LLMs via Imagination, Searching, and Criticizing18 Apr

Other references · 18

  1. 2025DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learningcited by 6,330
  2. 2022Self-Consistency Improves Chain of Thought Reasoning in Language Modelscited by 5,708
  3. 2023Sparse Autoencoders Find Highly Interpretable Features in Language Modelscited by 1,420
  4. 2024Tulu 3: Pushing Frontiers in Open Language Model Post-Trainingcited by 1,053
  5. 2025Does Reinforcement Learning Really Incentivize Reasoning Capacity in LLMs Beyond the Base Model?cited by 982
  6. 2023Plan-and-Solve Prompting: Improving Zero-Shot Chain-of-Thought Reasoning by Large Language Modelscited by 560
  7. 2024Jailbreaking Leading Safety-Aligned LLMs with Simple Adaptive Attackscited by 486
  8. 2024Mixture-of-Agents Enhances Large Language Model Capabilitiescited by 438
  9. 2024Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2cited by 394
  10. 2024Transcoders Find Interpretable LLM Feature Circuitscited by 200
  11. 2024LLMs Can't Plan, But Can Help Planning in LLM-Modulo Frameworkscited by 179
  12. 2023CodePlan: Repository-level Coding using LLMs and Planningcited by 140
6 more
  1. 2024Critique-out-Loud Reward Modelscited by 106
  2. 2024Group Robust Preference Optimization in Reward-free RLHFcited by 95
  3. 2024LLMs Still Can't Plan; Can LRMs? A Preliminary Evaluation of OpenAI's o1 on PlanBenchcited by 94
  4. 2023Have LLMs Advanced Enough? A Challenging Problem Solving Benchmark For Large Language Modelscited by 77
  5. 2025Optimizing Language Models for Inference Time Objectives using Reinforcement Learningcited by 53
  6. 2024Inductive or Deductive? Rethinking the Fundamental Reasoning Abilities of LLMscited by 29

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 2; 2026: 8.

10 papers.

  1. 2025GraphRAG-R1: Graph Retrieval-Augmented Generation with Process-Constrained Reinforcement Learningcited by 12
  2. 2026Towards a Mechanistic Understanding of Large Reasoning Models: A Survey of Training, Inference, and Failurescited by 6
  3. 2026Re2: Unlocking LLM Reasoning via Reinforcement Learning with Re-solvingcited by 6
  4. 2025Explore Briefly, Then Decide: Mitigating LLM Overthinking via Cumulative Entropy Regulationcited by 4
  5. 2026Limited Reasoning Space: The cage of long-horizon reasoning in LLMscited by 3
  6. 2026CLORE: Content-Level Optimization for Reasoning Efficiencycited by 0
  7. 2026Cognitive Demand Steering for Adaptive Meta-Reasoning in Large Language Modelscited by 0
  8. 2026Revisiting Complete Reasoning Traces for Post-Trainingcited by 0
  9. 2026Splicing Hints into Reasoning Traces: Learning from Demonstrator Failurescited by 0
  10. 2026Argument Graph Uncertainty: Quantifying Uncertainty from the Logical Structure of Reasoning Chainscited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026Thinking Past the Answer: Evaluating Harmful Overthinking in Large Reasoning Models
  2. 2025Stop Overthinking: A Survey on Efficient Reasoning for Large Language Models
  3. 2025Don't Think Longer, Think Wisely: Optimizing Thinking Dynamics for Large Reasoning Models
  4. 2025Large Language and Reasoning Models are Shallow Disjunctive Reasoners
  5. 2026To Think or Not To Think, That is The Question for Large Reasoning Models in Theory of Mind Tasks
  6. 2025The relationship between reasoning and performance in large language models--o3 (mini) thinks harder, not longer
  7. 2025Why Does Reasoning Length Converge? Unveiling the Underfitting-Overfitting Trade-off in Chain-of-Thought
  8. 2025Don't Overthink It: A Survey of Efficient R1-style Large Reasoning Models