mlsanityData to 14 Oct 2026

Explainable Chain-of-Thought Reasoning: An Empirical Analysis on State-Aware Reasoning Dynamics

arXiv 2509.00190 ↗First posted 29 Aug 2025cs.CL cs.AIcited by 16

On the map: LLM reasoning → Prompting and chain-of-thought reasoning

Abstract (arXiv)

Recent advances in chain-of-thought (CoT) prompting have enabled large language models (LLMs) to perform multi-step reasoning. However, the explainability of such reasoning remains limited, with prior work primarily focusing on local token-level attribution, such that the high-level semantic roles of reasoning steps and their transitions remain underexplored. In this paper, we introduce a state-aware transition framework that abstracts CoT trajectories into structured latent dynamics. Specifically, to capture the evolving semantics of CoT reasoning, each reasoning step is represented via spectral analysis of token-level embeddings and clustered into semantically coherent latent states. To characterize the global structure of reasoning, we model their progression as a Markov chain, yielding a structured and interpretable view of the reasoning process. This abstraction supports a range of analyses, including semantic role identification, temporal pattern visualization, and consistency evaluation.

Key facts. Cited by 16 papers in the map (data through 14 Oct 2026). Placed in Prompting and chain-of-thought reasoning, one of 17 directions in LLM reasoning. Builds on 21 of its 25 references that are in the map. First posted on arXiv 29 Aug 2025.

Builds on

21 of 25 references.

Closest ancestors · 5

Same direction, newest first

  1. 2025Unlocking General Long Chain-of-Thought Reasoning Capabilities of Large Language Models via Representation Engineering14 Mar
  2. 2024Training Large Language Models to Reason in a Continuous Latent Space9 Dec
  3. 2023Analyzing Chain-of-Thought Prompting in Large Language Models via Gradient-based Feature Attributions25 Jul
  4. 2023Language Models Don't Always Say What They Think: Unfaithful Explanations in Chain-of-Thought Prompting7 May
  5. 2022Large Language Models are Zero-Shot Reasoners24 May

Other references · 16

  1. 2021Training Verifiers to Solve Math Word Problemscited by 10,109
  2. 2021Measuring Mathematical Problem Solving With the MATH Datasetcited by 5,371
  3. 2018HotpotQA: A Dataset for Diverse, Explainable Multi-hop Question Answeringcited by 3,754
  4. 2024Gemma 2: Improving Open Language Models at a Practical Sizecited by 1,998
  5. 2018CommonsenseQA: A Question Answering Challenge Targeting Commonsense Knowledgecited by 1,822
  6. 2024SpinQuant: LLM quantization with learned rotationscited by 471
  7. 2024Lookback Lens: Detecting and Mitigating Contextual Hallucinations in Large Language Models Using Only Attention Mapscited by 142
  8. 2023Faithfulness Tests for Natural Language Explanationscited by 107
  9. 2023Towards a Mechanistic Interpretation of Multi-Step Reasoning Capabilities of Language Modelscited by 50
  10. 2023Does Physical Adversarial Example Really Matter to Autonomous Driving? Towards System-Level Effect of Adversarial Object Evasion Attackcited by 37
  11. 2025Rethinking External Slow-Thinking: From Snowball Errors to Probability of Correct Reasoningcited by 36
  12. 2023Towards Understanding In-Context Learning with Contrastive Demonstrations and Saliency Mapscited by 35
4 more
  1. 2024Towards Uncovering How Large Language Model Works: An Explainability Perspectivecited by 22
  2. 2025CTRLS: Chain-of-Thought Reasoning via Latent State-Transitioncited by 14
  3. 2025Beyond Words: A Latent Memory Approach to Internal Reasoning in LLMscited by 8
  4. 2024CogErgLLM: Exploring Large Language Model Systems Design Perspective Using Cognitive Ergonomicscited by 3

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 1; 2026: 15.

16 papers.

Sort:

By area:

  1. 2026WS-GRPO: Weakly-Supervised Group-Relative Policy Optimization for Rollout-Efficient Reasoningcited by 12
  2. 2026Generate, Filter, Control, Replay: A Comprehensive Survey of Rollout Strategies for LLM Reinforcement Learningcited by 9
  3. 2026Streaming Hallucination Detection in Long Chain-of-Thought Reasoningcited by 5
  4. 2026Unveiling the Entropy Dynamics of Chain-of-Thought Reasoningcited by 4
  5. 2026Stepwise Penalization for Length-Efficient Chain-of-Thought Reasoningcited by 3
  6. 2026Saliency-R1: Enforcing Interpretable and Faithful Vision-language Reasoning via Saliency-map Alignment Rewardcited by 2
  7. 2026OLIVIA: Online Learning via Inference-time Action Adaptation for Decision Making in LLM ReAct Agentscited by 2
  8. 2026ProSR: Process-Shaped Spatial Reasoning for Reliable Chain-of-Thought in VLMscited by 2
  9. 2026AutoCRAT: Within-trajectory Joint Control of Stochasticity and Compute for LLM Reasoningcited by 2
  10. 2026PRISM: A Dual View of LLM Reasoning through Semantic Flow and Latent Computationcited by 1
  11. 2026Skill-CMIB: Multimodal Agent Skill for Consistent Action via Conditional Multimodal Information Bottleneckcited by 1
  12. 2026F-GRPO: Factorized Group-Relative Policy Optimization for Unified Candidate Generation and Rankingcited by 1
4 more
  1. 2025Step-Tagging: Toward controlling the generation of Language Reasoning Models through step monitoringcited by 0
  2. 2026TRACES: Tagging Reasoning Steps for Adaptive Cost-Efficient Early-Stoppingcited by 0
  3. 2026Interpreting Latent CoT Reasoning as Dynamical Systemscited by 0
  4. 2026CoT-Core: Accelerating LLM Evaluation via CoT-Aware Coreset Selectioncited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026When does Chain-of-Thought Help: A Markovian Perspective
  2. 2025Reasoning Beyond Language: A Comprehensive Survey on Latent Chain-of-Thought Reasoning
  3. 2024How to think step-by-step: A mechanistic understanding of chain-of-thought reasoning
  4. 2024Enhancing Chain of Thought Prompting in Large Language Models via Reasoning Patterns
  5. 2026How Far Ahead Do LLMs Plan? Uncovering the Latent Horizon in Chain-of-Thought Reasoning
  6. 2026J-CoT: Chain-of-Thought in J-Space
  7. 2022Towards Understanding Chain-of-Thought Prompting: An Empirical Study of What Matters
  8. 2026Mechanistic Interpretability of Chain-of-Thought Reasoning via Sequential Activation Patching