Think and Answer ME: Benchmarking and Exploring Multi-Entity Reasoning Grounding in Remote Sensing

Shuchang Lyu, Haiquan Wen, Guangliang Cheng +5

Recent advances in reasoning language models and reinforcement learning with verifiable rewards have significantly enhanced multi-step reasoning capabilities. This progress motivat…