mlsanityData to 7 Oct 2026

Circular Reasoning: Understanding Self-Reinforcing Loops in Large Reasoning Models

arXiv 2601.05693 ↗First posted 9 Jan 2026cs.AIcited by 5

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

Abstract (arXiv)

Despite the success of test-time scaling, Large Reasoning Models (LRMs) frequently encounter repetitive loops that lead to computational waste and inference failure. In this paper, we identify a distinct failure mode termed Circular Reasoning. Unlike traditional model degeneration, this phenomenon manifests as a self-reinforcing trap where generated content acts as a logical premise for its own recurrence, compelling the reiteration of preceding text. To systematically analyze this phenomenon, we introduce LoopBench, a dataset designed to capture two distinct loop typologies: numerical loops and statement loops. Mechanistically, we characterize circular reasoning as a state collapse exhibiting distinct boundaries, where semantic repetition precedes textual repetition. We reveal that reasoning impasses trigger the loop onset, which subsequently persists as an inescapable cycle driven by a self-reinforcing V-shaped attention mechanism. Guided by these findings, we employ the Cumulative Sum (CUSUM) algorithm to capture these precursors for early loop prediction. Experiments across diverse LRMs validate its accuracy and elucidate the stability of long-chain reasoning.

Key facts. Cited by 5 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 18 of its 39 references that are in the map. First posted on arXiv 9 Jan 2026.

Builds on

18 of 39 references.

Closest ancestors · 5

Same direction, newest first

  1. 2025Word Salad Chopper: Reasoning Models Waste A Ton Of Decoding Budget On Useless Repetitions, Self-Knowingly1 Nov
  2. 2025LLM Latent Reasoning as Chain of Superposition17 Oct
  3. 2025The Evolution of Thought: Tracking LLM Overthinking via Reasoning Dynamics Analysis25 Aug
  4. 2025Missing Premise exacerbates Overthinking: Are Reasoning Models losing Critical Thinking Skill?9 Apr
  5. 2024Scaling LLM Test-Time Compute Optimally can be More Effective than Scaling Model Parameters6 Aug

Other references · 13

  1. 2025DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learningcited by 6,146
  2. 2019CTRL: A Conditional Transformer Language Model for Controllable Generationcited by 754
  3. 2023SemDeDup: Data-efficient learning at web-scale through semantic deduplicationcited by 295
  4. 2023Function Vectors in Large Language Modelscited by 292
  5. 2022Scaling Laws and Interpretability of Learning from Repeated Datacited by 147
  6. 2025SuperGPQA: Scaling LLM Evaluation across 285 Graduate Disciplinescited by 67
  7. 2024Knowledge Circuits in Pretrained Transformerscited by 60
  8. 2025Innate Reasoning is Not Enough: In-Context Learning Enhances Reasoning Large Language Models with Less Overthinkingcited by 19
  9. 2024Everything is Editable: Extend Knowledge Editing to Unstructured Data in Large Language Modelscited by 18
  10. 2025Related Knowledge Perturbation Matters: Rethinking Multiple Pieces of Knowledge Editing in Same-Subjectcited by 10
  11. 2025Focusing by Contrastive Attention: Enhancing VLMs' Visual Reasoningcited by 8
  12. 2025Large Language Model Sourcing: A Surveycited by 5
1 more
  1. 2025GloSS over Toxicity: Understanding and Mitigating Toxicity in LLMs via Global Toxic Subspacecited by 4

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 5.

5 papers.

  1. 2026Latent-GRPO: Group Relative Policy Optimization for Latent Reasoningcited by 10
  2. 2026ShortOPD: Recovering Pruned LLMs with Short-to-Long On-Policy Distillationcited by 7
  3. 2026CUSUM-Shaped Inference-Time Monitoring and Targeted Re-Decoding for Quantized Small Language Model Reasoningcited by 1
  4. 2026Sanity Checks for Long-Form Hallucination Detectioncited by 0
  5. 2026Calibrated e-CUSUM Decoding for Quantized Reasoning Models: Why Token Log-Probability Is the Wrong Observable for Decoding Monitorscited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026A Mechanistic Analysis of Looped Reasoning Language Models
  2. 2026Stop When Reasoning Converges: Semantic-Preserving Early Exit for Reasoning Models
  3. 2026Can We Break LLMs Out of Self-Loops? Fine-Grained Reasoning Control with Activation Steering
  4. 2026Dissecting Failure Dynamics in Large Language Model Reasoning
  5. 2026Beyond the Commitment Boundary: Probing Epiphenomenal Chain-of-Thought in Large Reasoning Models
  6. 2025Lost at the Beginning of Reasoning
  7. 2026Large Language Models Decide Early and Explain Later
  8. 2025The Illusion of Thinking: Understanding the Strengths and Limitations of Reasoning Models via the Lens of Problem Complexity