Fetching the paper…

Steve-Eye: Equipping LLM-based Embodied Agents with Visual Perception in Open Worlds · Around