Visual thinking should not only sound right; it should show its evidence. While recent vision-language models (VLMs) can produce natural-language reasoning traces, these traces often leave the supporting image regions implicit, making them hard to verify and difficult to supervise. We introduce visually grounded thinking, a reasoning process in which models interleave natural-language thoughts with explicit point or box groundings of the visual evidence used at each step. This lets the model express intermediate reasoning in language while grounding key objects in the image regions they refer to. To train this behavior, we construct a scalable synthesis pipeline that distills correct visual reasoning traces, extracts the visual objects required by the traces, grounds them with a SAM3-based agent, and derives aligned point and box supervision from the resulting masks. We further propose grounding-aware reinforcement learning, which combines answer correctness rewards with dense grounding rewards that score whether generated object references match the correct image evidence. Across two counting benchmarks and four spatial reasoning benchmarks, adding visually grounded thinking to Gemma3-4B-IT consistently improves performance over the original model and the non-grounded thinking baseline. On spatial reasoning, the visually grounded thinking 4B models match, and in some cases surpass, Gemma3-27B-IT from the same model family. Our analysis shows that point grounding is well suited to counting, while box grounding benefits most from explicit grounding rewards on spatial tasks. Overall, our results show that VLMs think better when their intermediate thoughts are tied to the image regions that make them true.

翻译：视觉推理不仅应当言之有理，更需呈现其证据。尽管当前的视觉语言模型（VLM）能够生成自然语言的推理链，但这些推理链往往将所依托的图像区域隐含其中，导致难以验证且不便监督。本文提出视觉锚定推理（visually grounded thinking）——一种推理过程，使模型能在自然语言思维中交替嵌入显式的点或框标注，以标示每一步所用到的视觉证据。这使模型既能用语言表达中间推理，又能将关键对象锚定于其指代的图像区域。为训练此行为，我们构建了可扩展的合成流程：提炼正确的视觉推理轨迹，提取轨迹所需的视觉对象，借助基于SAM3的智能体对其锚定，并从生成的掩码中导出对齐的点与框监督信号。我们进一步提出锚定感知强化学习（grounding-aware reinforcement learning），将答案正确性奖励与密集的锚定奖励相结合，后者用于评估生成的对象引用是否匹配正确的图像证据。在两个计数基准与四个空间推理基准上，为Gemma3-4B-IT模型添加视觉锚定推理，其性能持续优于原始模型及无锚定推理基线。在空间推理任务中，视觉锚定推理的4B模型已达到甚至超越同系列Gemma3-27B-IT模型的表现。分析表明，点锚定更适用于计数任务，而框锚定则在空间任务中从显式锚定奖励中获益最多。总体而言，我们的结果表明，当VLMs的中间思维与其所依凭的图像区域紧密关联时，其推理质量会显著提升。