paper-with-me

Papers

Statistical Early Stopping for Reasoning Models

2026-02-15 · Yangxinyu Xie, Tao Wang, Soham Mallick, Yan Sun, Georgy Noarov, Mengxin Yu, Tanwi Mallick, Weijie J. Su, Edgar Dobriban arxiv

While LLMs have seen substantial improvement in reasoning capabilities, they also sometimes overthink, generating unnecessary reasoning steps, particularly under uncertainty, given ill-posed or ambiguous queries. We introduce statistically principled early stopping methods that monitor uncertainty signals during generation to mitigate this issue. Our first approach is parametric: it models inter-arrival times of uncertainty keywords as a renewal process and applies sequential testing for stopping. Our second approach is nonparametric and provides finite-sample guarantees on the probability of halting too early on well-posed queries. We conduct empirical evaluations on reasoning tasks across several domains and models. Our results indicate that uncertainty-aware early stopping can improve both efficiency and reliability in LLM reasoning, and we observe especially significant gains for math reasoning.

📄 PDF Abstract BibTeX arXiv:2602.13935

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

TRACES: Tagging Reasoning Steps for Adaptive Cost-Efficient Early-Stopping

2026-04-22 · Yannis Belkhiter, Seshu Tirupathi, Giulio Zizzo, John D. Kelleher arxiv

The field of Language Reasoning Models (LRMs) has been very active over the past few years with advances in training and inference techniques enabling LRMs to reason longer, and more accurately. However, a growing body o…

Mathematical Reasoning

Don't stop me now: Rethinking Validation Criteria for Model Parameter Selection

2026-02-25 · Andrea Apicella, Francesco Isgrò, Andrea Pollastro, Roberto Prevete arxiv

Despite the extensive literature on training loss functions, the evaluation of generalization on the validation set remains underexplored. In this work, we conduct a systematic empirical and statistical study of how the …

ESTAR: Early-Stopping Token-Aware Reasoning For Efficient Inference

2026-02-10 · Junda Wang, Zhichao Yang, Dongxu Zhang, Sanjit Singh Batra 외 arxiv

Large reasoning models (LRMs) achieve state-of-the-art performance by generating long chains-of-thought, but often waste computation on redundant reasoning after the correct answer has already been reached. We introduce …

Reinforcement LearningDomain Generalization

NYTRO: When Subsampling Meets Early Stopping

2015-10-19 · Tomas Angles, Raffaello Camoriano, Alessandro Rudi, Lorenzo Rosasco

Early stopping is a well known approach to reduce the time complexity for performing training and model selection of large scale learning machines. On the other hand, memory/space (rather than time) complexity is the mai…

Model Selectionregression

Early Stopping for Large Reasoning Models via Confidence Dynamics

2026-04-06 · Parsa Hosseini, Sumit Nawathe, Mahdi Salmani, Meisam Razaviyayn 외 arxiv

Large reasoning models rely on long chain-of-thought generation to solve complex problems, but extended reasoning often incurs substantial computational cost and can even degrade performance due to overthinking. A key ch…