Reinforcement learning with verifiable rewards (RLVR) has advanced reasoning capabilities in multimodal large language models. However, existing methods typically treat visual inputs as deterministic, overlooking the perceptual ambiguity inherent to the visual modality. Consequently, they fail to distinguish whether a model's uncertainty stems from complex reasoning or ambiguous perception, preventing the targeted allocation of exploration or learning signals. To address this gap, we introduce \textbf{DUPL}, a dual-uncertainty guided policy learning approach for multimodal RLVR that quantifies and leverages both perceptual uncertainty (via symmetric KL divergence) and output uncertainty (via policy entropy) to guide policy updates. By establishing an uncertainty-driven feedback loop and employing a dynamic branch prioritization mechanism, DUPL recalibrates the policy advantage to focus learning on states with high perceptual or decisional ambiguity, enabling effective targeted exploration beyond passive data augmentation. Evaluated on diverse multimodal reasoning benchmarks spanning mathematical and general domains, DUPL achieves solid gains. It improves Qwen2.5-VL accuracy by up to $\textbf{12.3%}$ (3B) and $\textbf{7.9%}$ (7B), and Qwen3-VL-Instruct by up to $\textbf{10.7%}$ (4B) and $\textbf{12.4%}$ (8B), consistently outperforming GRPO, while seamlessly generalizing to alternative algorithms (DAPO, $\textbf{+6.5%}$ avg) and architectures (LLaVA-OneVision-1.5, $\textbf{+4.7%}$ avg). These results demonstrate that DUPL is an effective and generalizable approach for multimodal RLVR.

翻译：基于可验证奖励的强化学习（RLVR）提升了多模态大语言模型的推理能力。然而，现有方法通常将视觉输入视为确定性信息，忽视了视觉模态固有的感知模糊性。因此，它们无法区分模型的不确定性来源于复杂推理还是模糊感知，从而无法定向分配探索或学习信号。为解决这一空白，我们提出**DUPL**——一种用于多模态RLVR的双不确定性引导策略学习方法，该方法通过量化并利用感知不确定性（基于对称KL散度）和输出不确定性（基于策略熵）来引导策略更新。通过建立不确定性驱动的反馈循环并采用动态分支优先化机制，DUPL重新校准策略优势，将学习聚焦于具有高感知或决策模糊性的状态，从而在被动数据增强之外实现有效的定向探索。在涵盖数学与通用领域的多模态推理基准测试中，DUPL取得了显著提升。它将Qwen2.5-VL的准确率提升了最高**12.3%**（3B模型）和**7.9%**（7B模型），将Qwen3-VL-Instruct提升了最高**10.7%**（4B模型）和**12.4%**（8B模型），持续超越GRPO，同时可无缝推广至其他算法（DAPO，平均**+6.5%**）和架构（LLaVA-OneVision-1.5，平均**+4.7%**）。这些结果表明，DUPL是一种有效且可泛化的多模态RLVR方法。