GEM: Generative Supervision Helps Embodied Intelligence 事件

GEM: Generative Supervision Helps Embodied Intelligence arXiv:2605.28548v1 Announce Type: new Abstract: Embodied Vision-Language Models (VLMs) have demonstrated impressive performance and generalization in robotics, particularly within Vision-Language-Action frameworks. However, a significant gap remains between the high-level semantic focus of standard text-guided pre-training paradigms and the low-level spatial and physical knowledge critical for execution in embodied environments. In this pa