mlsanityData to 14 Oct 2026

Entropy trajectory shape predicts LLM reasoning reliability: A diagnostic study of uncertainty dynamics in chain-of-thought

arXiv 2603.18940 ↗First posted 19 Mar 2026cs.CL cs.LGcited by 8

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

Abstract (arXiv)

Understanding uncertainty in chain-of-thought reasoning is critical for reliable deployment of large language models. In this work, we propose a simple yet effective diagnostic approach based on trajectory shape rather than scalar magnitude. We show that this signal is practical, interpretable, and inexpensive to obtain in black-box settings, while remaining robust across models and datasets. Through extensive ablations and cross-domain replications, we demonstrate its utility for selective prediction and triage. Our findings offer a generalizable insight into uncertainty dynamics in reasoning tasks, with particular focus on numeric and discrete-answer settings.

Key facts. Cited by 8 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 15 of its 26 references that are in the map. First posted on arXiv 19 Mar 2026.

Builds on

15 of 26 references.

Closest ancestors · 7

Same direction, newest first

  1. 2026Learning When to Sample: Confidence-Aware Selective Sampling for Efficient Chain-of-Thought Reasoning9 Mar
  2. 2026ConfSpec: Efficient Step-Level Speculative Reasoning via Confidence-Gated Verification28 Jan
  3. 2026Recurrent Confidence Chain: Temporal-Aware Uncertainty Quantification in Large Language Models19 Jan
  4. 2025Temporalizing Confidence: Evaluation of Chain-of-Thought Reasoning with Signal Temporal Logic9 Jun
  5. 2025Thought calibration: Efficient and confident test-time scaling23 May
  6. 2023Universal Self-Consistency for Large Language Model Generation29 Nov
  7. 2022Language Models (Mostly) Know What They Know11 Jul

Other references · 8

  1. 2021Training Verifiers to Solve Math Word Problemscited by 10,109
  2. 2024Qwen2.5 Technical Reportcited by 4,942
  3. 2023Let's Verify Step by Stepcited by 3,789
  4. 2022Solving math word problems with process- and outcome-based feedbackcited by 901
  5. 2023Active Prompting with Chain-of-Thought for Large Language Modelscited by 154
  6. 2025Can LLMs Predict Their Own Failures? Self-Awareness via Internal Circuitscited by 14
  7. 2026VI-CuRL: Stabilizing Verifier-Independent RL Reasoning via Confidence-Guided Variance Reductioncited by 10
  8. 2026Know When You're Wrong: Aligning Confidence with Correctness for LLM Error Detectioncited by 8

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 8.

8 papers.

  1. 2026When Do LLMs Reason? A Dynamical Systems View via Entropy Phase Transitionscited by 2
  2. 2026ProSR: Process-Shaped Spatial Reasoning for Reliable Chain-of-Thought in VLMscited by 2
  3. 2026How Language Models Fail: Token-Level Signatures of Committed and Persistent Reasoning Failurescited by 1
  4. 2026The Tell-Tale Trace: Detecting Reasoning Failures in LLMs Using Chain-of-Thought Dynamicscited by 1
  5. 2026How Hard Does It Think? Analyzing Step-Aware Reasoning Energy in LLM Chain-of-Thought Trajectoriescited by 0
  6. 2026Chain-of-Thought Entropy as a Reliability Signal: A Preregistered Reproductioncited by 0
  7. 2026Learning More by Training Less: Uncertainty-Guided Progressive Contraction for Efficient Reasoning Post-Trainingcited by 0
  8. 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. 2026Answer-Distribution Trajectories: A Stochastic-Dynamics View of LLM Reasoning
  2. 2026EDIS: Diagnosing LLM Reasoning via Entropy Dynamics
  3. 2026Unveiling the Entropy Dynamics of Chain-of-Thought Reasoning
  4. 2026Beyond Scalars: Evaluating and Understanding LLM Reasoning via Geometric Progress and Stability
  5. 2026ETR: Entropy Trend Reward for Efficient Chain-of-Thought Reasoning
  6. 2026How Far Ahead Do LLMs Plan? Uncovering the Latent Horizon in Chain-of-Thought Reasoning
  7. 2026Reasoning Models Don't Just Think Longer, They Move Differently
  8. 2026LLM Reasoning as Trajectories: Step-Specific Representation Geometry and Correctness Signals