让视觉语言模型像鸽子一样主动探索，提升空间推理能力。
Active Exploring like a Pigeon: Reinforcing Spatial Reasoning via Agentic Vision-Language Models
- 用动态认知地图记录物体位置方向，实现持续记忆。
- 通过空间陈述代码生成密集奖励，提升复杂任务表现。
- 适合需要主动感知与空间规划的智能体研究者。
让视觉语言模型（VLMs）具备空间推理能力仍具挑战。现有方法将VLM视为被动观察者，难以应用于真实场景；而强化学习依赖稀疏奖励，限制了复杂推理任务的效果。受鸽子构建和利用认知地图导航的启发，我们提出一种新型代理式空间推理框架。首先引入动态认知地图，以物体位置和朝向参数化场景布局，作为新观测的持久记忆。其次提出空间陈述代码（SAC），用Python表达式程序化描述空间关系。结合动态认知地图，SAC可验证中间推理步骤，提供密集奖励信号。通过监督与强化微调优化模型。在MindCube基准测试中，整体准确率达80.5%，在具有挑战性的旋转子集上优于当前最佳方法29.5个百分点（相对提升53.2%）。代码与数据已开源。
原文摘要 · Abstract (English)
Enabling Vision-Language Models (VLMs) to perform spatial reasoning remains challenging. Existing approaches treat VLMs as passive observers, which is difficult for real-world applications. Moreover, reinforcement learning methods rely on sparse rewards, limiting their effectiveness for complex reasoning tasks. Inspired by pigeons' building and exploiting cognitive maps for navigation, we propose a novel agentic pipeline for spatial reasoning. First, we introduce a new \emph{dynamic cognitive map} parameterizing scene layout as object positions and orientations, serving as persistent memory for new observations. Second, we propose a novel \emph{Spatial Assertion Codes (SAC)}, Python expressions programmatically describing spatial relationships. By collaborating with the dynamic cognitive map, SAC enables verification of intermediate reasoning steps, providing dense reward signals. We optimize the model via supervised and reinforcement finetuning. Experiments on the MindCube benchmark demonstrate state-of-the-art performance with \emph{80.5\%} overall accuracy, outperforming the best current method by \emph{29.5} accuracy points (a relative improvement of \emph{53.2\%}) on the challenging \textsc{Rotation} subset. Our code and data are open-sourced at https://github.com/dw-dengwei/active-spatial-reasoning.git.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。