Image de couverture de livre

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

Fuente: arXiv

Enregistré dans:

Détails bibliographiques
Auteur principal:	Zhao, Xinghao
Format:	Preprint
Publié:	2026
Sujets:	Computation and Language Machine Learning
Accès en ligne:	Acceder al recurso
Tags:	Ajouter un tag Pas de tags, Soyez le premier à ajouter un tag!

Documents similaires

How to think step-by-step: A mechanistic understanding of chain-of-thought reasoning
par: Dutta, Subhabrata, et autres
Publié: (2024)

How well can off-the-shelf LLMs elucidate molecular structures from mass spectra using chain-of-thought reasoning?
par: Wang, Yufeng, et autres
Publié: (2026)

To CoT or not to CoT? Chain-of-thought helps mainly on math and symbolic reasoning
par: Sprague, Zayne, et autres
Publié: (2024)

Your thoughts tell who you are: Characterize the reasoning patterns of LRMs
par: Chen, Yida, et autres
Publié: (2025)

Encode, Think, Decode: Scaling test-time reasoning with recursive latent thoughts
par: Koishekenov, Yeskendir, et autres
Publié: (2025)

Measuring and curing reasoning rigidity: from decorative chain-of-thought to genuine faithfulness
par: Basu, Abhinaba, et autres
Publié: (2026)

Large language models can learn and generalize steganographic chain-of-thought under process supervision
par: Skaf, Joey, et autres
Publié: (2025)

Activation Control for Efficiently Eliciting Long Chain-of-thought Ability of Language Models
par: Zhao, Zekai, et autres
Publié: (2025)

Probabilistic unifying relations for modelling epistemic and aleatoric uncertainty: semantics and automated reasoning with theorem proving
par: Ye, Kangfeng, et autres
Publié: (2023)

EPO: Entropy-regularized Policy Optimization for LLM Agents Reinforcement Learning
par: Xu, Wujiang, et autres
Publié: (2025)

AI-generated data contamination erodes pathological variability and diagnostic reliability
par: He, Hongyu, et autres
Publié: (2026)

OCEAN: Offline Chain-of-thought Evaluation and Alignment in Large Language Models
par: Wu, Junda, et autres
Publié: (2024)

When an LLM is apprehensive about its answers -- and when its uncertainty is justified
par: Sychev, Petr, et autres
Publié: (2025)

Entropy Law: The Story Behind Data Compression and LLM Performance
par: Yin, Mingjia, et autres
Publié: (2024)

On multi-token prediction for efficient LLM inference
par: Mehra, Somesh, et autres
Publié: (2025)

Can Language Models Perform Robust Reasoning in Chain-of-thought Prompting with Noisy Rationales?
par: Zhou, Zhanke, et autres
Publié: (2024)

Sudoku-Bench: Evaluating creative reasoning with Sudoku variants
par: Seely, Jeffrey, et autres
Publié: (2025)

Addressing Performance Saturation for LLM RL via Precise Entropy Curve Control
par: Li, Bolian, et autres
Publié: (2026)

Harnessing Uncertainty: Entropy-Modulated Policy Gradients for Long-Horizon LLM Agents
par: Wang, Jiawei, et autres
Publié: (2025)

Beyond Semantic Entropy: Boosting LLM Uncertainty Quantification with Pairwise Semantic Similarity
par: Nguyen, Dang, et autres
Publié: (2025)

Language models scale reliably with over-training and on downstream tasks
par: Gadre, Samir Yitzhak, et autres
Publié: (2024)

Boosting classification reliability of NLP transformer models in the long run
par: Kmetty, Zoltán, et autres
Publié: (2023)

Intrinsic Entropy of Context Length Scaling in LLMs
par: Shi, Jingzhe, et autres
Publié: (2025)

Language hooks: a modular framework for augmenting LLM reasoning that decouples tool usage from the model and its prompt
par: de Mijolla, Damien, et autres
Publié: (2024)

When can transformers reason with abstract symbols?
par: Boix-Adsera, Enric, et autres
Publié: (2023)

Artificial Expert Intelligence through PAC-reasoning
par: Shalev-Shwartz, Shai, et autres
Publié: (2024)

Navigation under uncertainty: Trajectory prediction and occlusion reasoning with switching dynamical systems
par: Wei, Ran, et autres
Publié: (2024)

Density estimation with LLMs: a geometric investigation of in-context learning trajectories
par: Liu, Toni J. B., et autres
Publié: (2024)

Is continuous CoT better suited for multi-lingual reasoning?
par: Bashir, Ali Hamza, et autres
Publié: (2026)

Are complicated loss functions necessary for teaching LLMs to reason?
par: Carrino, Gabriele, et autres
Publié: (2026)

Statistical multi-metric evaluation and visualization of LLM system predictive performance
par: Ackerman, Samuel, et autres
Publié: (2025)

Few shot chain-of-thought driven reasoning to prompt LLMs for open ended medical question answering
par: Nachane, Saeel Sandeep, et autres
Publié: (2024)

AutoRule: Reasoning Chain-of-thought Extracted Rule-based Rewards Improve Preference Learning
par: Wang, Tevin, et autres
Publié: (2025)

Neural networks for abstraction and reasoning: Towards broad generalization in machines
par: Bober-Irizar, Mikel, et autres
Publié: (2024)

Functional Entropy: Predicting Functional Correctness in LLM-Generated Code with Uncertainty Quantification
par: Bouchard, Dylan, et autres
Publié: (2026)

Explore Briefly, Then Decide: Mitigating LLM Overthinking via Cumulative Entropy Regulation
par: Bin, Yi, et autres
Publié: (2025)

LANCET: Neural Intervention via Structural Entropy for Mitigating Faithfulness Hallucinations in LLMs
par: Wang, Chenxu, et autres
Publié: (2026)

Entropy-Regularized Process Reward Model
par: Zhang, Hanning, et autres
Publié: (2024)

Failure Modes of Maximum Entropy RLHF
par: Çağatan, Ömer Veysel, et autres
Publié: (2025)

Entropy Polarity in Reinforcement Fine-Tuning: Direction, Asymmetry, and Control
par: Zhang, Jiazheng, et autres
Publié: (2026)