GeoChain：用于地理推理的多模态思维链
GeoChain: Multimodal Chain-of-Thought for Geographic Reasoning
AI总结：
GeoChain是一个大规模基准，通过146万张街景图像和21步思维链问题序列，评估多模态大语言模型在视觉、空间、文化和精确地理位置四类推理中的逐步地理推理能力，揭示模型在视觉基础和准确定位上的持续挑战。
摘要：
本文介绍了GeoChain，这是一个用于评估多模态大语言模型（MLLMs）中逐步地理推理能力的大规模基准。利用146万张Mapillary街景图像，GeoChain将每张图像与一个包含21步思维链（CoT）的问题序列配对（共计超过3000万个问答对）。这些序列引导模型从粗粒度属性逐步过渡到细粒度定位，涵盖四个推理类别——视觉、空间、文化和精确地理位置——并按难度进行标注。图像还补充了语义分割（150个类别）和视觉可定位性评分。我们在包含2088张图像的多样化子集上对当代MLLMs（GPT-4.1系列、Claude 3.7、Gemini 2.5系列）进行了基准测试，结果揭示了持续存在的挑战：模型经常在视觉基础方面表现出弱点，显示出不稳定的推理，并且难以实现准确定位，尤其是当推理复杂性增加时。GeoChain提供了一种稳健的诊断方法，对于推动MLLMs在复杂地理推理方面取得重大进展至关重要。
英文摘要：
This paper introduces GeoChain, a large-scale benchmark for evaluating step-by-step geographic reasoning in multimodal large language models (MLLMs). Leveraging 1.46 million Mapillary street-level images, GeoChain pairs each image with a 21-step chain-of-thought (CoT) question sequence (over 30 million Q&A pairs). These sequences guide models from coarse attributes to fine-grained localization across four reasoning categories - visual, spatial, cultural, and precise geolocation - annotated by difficulty. Images are also enriched with semantic segmentation (150 classes) and a visual locatability score. Our benchmarking of contemporary MLLMs (GPT-4.1 variants, Claude 3.7, Gemini 2.5 variants) on a diverse 2,088-image subset reveals consistent challenges: models frequently exhibit weaknesses in visual grounding, display erratic reasoning, and struggle to achieve accurate localization, especially as the reasoning complexity escalates. GeoChain offers a robust diagnostic methodology, critical for fostering significant advancements in complex geographic reasoning within MLLMs.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
