Recognition through Reasoning: Reinforcing Image Geo-localization with Large Vision-Language Models | Read Paper on Bytez