Reinforcement Learning (RL) has significantly improved large language model reasoning, but existing RL fine-tuning methods rely heavily on heuristic techniques such as entropy regularization and reweighting to maintain stability. In practice, they often experience late-stage performance collapse, leading to degraded reasoning quality and unstable training. We derive that the magnitude of token-wise policy gradients in RL is negatively correlated with token probability and local policy entropy. Building on this result, we prove that training instability is driven by a tiny fraction of tokens, approximately 0.01\%, which we term \emph{spurious tokens}. When such tokens appear in correct responses, they contribute little to the reasoning outcome but inherit the full sequence-level reward, leading to abnormally amplified gradient updates. Motivated by this observation, we propose Spurious-Token-Aware Policy Optimization (STAPO) for large-scale model refining, which selectively masks such updates and renormalizes the loss over valid tokens. Across six mathematical reasoning benchmarks using Qwen 1.7B, 8B, and 14B base models, STAPO consistently demonstrates superior entropy stability and achieves an average performance improvement of 7.13\% over GRPO, 20-Entropy and JustRL.

翻译：强化学习（RL）显著提升了大语言模型的推理能力，但现有的RL微调方法严重依赖启发式技术（如熵正则化和权重调整）来维持稳定性。在实践中，这些方法常遭遇后期性能崩溃，导致推理质量下降和训练不稳定。我们推导得出，RL中基于标记的策略梯度幅度与标记概率及局部策略熵呈负相关。基于此结果，我们证明训练不稳定性由极少数（约0.01%）的标记驱动，我们称之为“伪标记”。当此类标记出现在正确响应中时，它们对推理结果的贡献微乎其微，却继承了完整的序列级奖励，导致梯度更新异常放大。受此观察启发，我们提出面向大规模模型精炼的伪标记感知策略优化（STAPO），该方法选择性屏蔽此类更新并对有效标记的损失进行重归一化。在使用Qwen 1.7B、8B和14B基础模型的六个数学推理基准测试中，STAPO始终展现出更优的熵稳定性，相比GRPO、20-Entropy和JustRL平均性能提升达7.13%。