GeoEvolve: Recursive Skill Evolution for Image Geolocalization

Image geolocalization with large vision-language models (LVLMs) can benefit from external evidence gathered through tools when visual reasoning alone is insufficient to resolve geographic ambiguity. However, unconstrained tool use often introduces distracting or conflicting signals. We introduce $\mathbf{GeoEvolve}$, a framework that recursively evolves $\text{GeoSkill}$, a single geolocalization skill encoding an explicit orchestration policy. $\mathbf{GeoEvolve}$ induces an initial policy from cold-start trajectories collected under minimally constrained tool use, and then enters an on-policy loop that generates new trajectories under its current guidance. At each iteration, $\mathbf{GeoEvolve}$ assigns process-level credit to effective and harmful decisions within a trajectory batch, and uses this feedback to preserve, clarify, or revise rules for hypothesis formation, tool selection, evidence verification, and stopping. On *Geo-EVO-Test* and *Im2GPS3K*, $\text{GeoSkill}$ consistently improves over unconstrained tool use and remains effective across LVLMs of different scales and families. Distilling skill-guided trajectories into a task-specific model yields further gains and sets a new state-of-the-art among compact models in image geolocalization. The code, data, and skills will be publicly released.