GeoAgent：借助强化地理特征学习在任意地点进行地理定位
GeoAgent: Learning to Geolocate Everywhere with Reinforced Geographic Characteristics
把地理相似性奖励与一致性奖励引入强化学习训练，为视觉语言模型在细粒度地理定位任务上提供可复用的奖励设计与专家标注思维链数据构建思路。
原文依据 we propose a geo-similarity reward and a consistency reward assessed by a consistency agent to assist training
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论本文提出 GeoAgent，一个能够与人类紧密推理并得出细粒度地址结论的模型。此前基于强化学习的方法在性能与可解释性上取得了突破，但由于依赖 AI 生成的思维链（CoT）数据与训练策略，而这些与地理特征相冲突，因此仍存在隐忧。为解决这些问题，我们首先引入 GeoSeek，一个由地理专家与专业玩家标注思维链数据的新地理定位数据集。我们进一步深入探究地理任务的固有特性，提出地理相似性奖励以及由一致性智能体评估的一致性奖励来辅助训练。这促使模型从地理视角向正确答案收敛，同时保证其推理过程的完整性与一致性。实验结果表明，GeoAgent 在多个粒度上优于现有方法与一系列通用视觉大语言模型，同时生成的推理与人类高度一致。预训练模型与数据将公开提供。
英文原文摘要 1131 字符
This paper presents GeoAgent, a model capable of reasoning closely with humans and deriving fine-grained address conclusions. Previous RL-based methods have achieved breakthroughs in performance and interpretability but still remain concerns because of their reliance on AI-generated chain-of-thought (CoT) data and training strategies, which conflict with geographic characteristics. To address these issues, we first introduce GeoSeek, a new geolocation dataset comprising CoT data annotated by geographic experts and professional players. We further thoroughly explore the inherent characteristics of geographic tasks and propose a geo-similarity reward and a consistency reward assessed by a consistency agent to assist training. This encourages the model to converge towards correct answers from a geographic perspective while ensuring the integrity and consistency of its reasoning process. Experimental results show that GeoAgent outperforms existing methods and a series of general VLLMs across multiple grains, while generating reasoning that closely aligns with humans. Pretrained model and data will be openly available.