BEAR: Benchmarking and Enhancing Multimodal Language Models for Atomic Embodied Capabilities

Open in new window