Large Reasoning Models (LRMs) often suffer from \emph{overthinking}, a phenomenon in which redundant reasoning steps are generated after a correct solution has already been reached. Existing early reasoning exit methods primarily rely on output-level heuristics or trained probing models to skip redundant reasoning steps, thereby mitigating overthinking. However, these approaches typically require additional rollout computation or externally labeled datasets. In this paper, we propose \textbf{NEAT}, a \textbf{N}euron-based \textbf{E}arly re\textbf{A}soning exi\textbf{T} framework that monitors neuron-level activation dynamics to enable training-free early exits, without introducing additional test-time computation. NEAT identifies exit-associated neurons and tracks their activation patterns during reasoning to dynamically trigger early exit or suppress reflection, thereby reducing unnecessary reasoning while preserving solution quality. Experiments on four reasoning benchmarks across six models with different scales and architectures show that, for each model, NEAT achieves an average token reduction of 22\% to 28\% when averaged over the four benchmarks, while maintaining accuracy.

翻译：大型推理模型（LRMs）常受“过度思考”现象困扰，即在已获得正确解后仍生成冗余推理步骤。现有早期推理退出方法主要依赖输出层启发式规则或训练后的探测模型来跳过冗余推理步骤，从而缓解过度思考。然而，这些方法通常需要额外的展开计算或外部标注数据集。本文提出\textbf{NEAT}，一种\textbf{基于神经元}的\textbf{早期推理退出}框架，通过监测神经元层面的激活动态实现无需训练的早期退出，且不引入额外的测试时计算开销。NEAT通过识别与退出相关的神经元，并在推理过程中追踪其激活模式，动态触发早期退出或抑制反思过程，从而在保持求解质量的同时减少不必要的推理。在涵盖六种不同规模与架构模型的四个推理基准测试中，NEAT对每个模型在四个基准上的平均标记减少量达到22\%至28\%，同时维持了准确率。