Embodied AI TopConf · NeuIPS2025

PRIMT: Preference-based Reinforcement Learning with Multimodal Feedback and Trajectory Synthesis from Foundation Models

NeuIPS2025 / Vision-Language-Action Model

视觉语言动作策略学习触觉感知
新标签打开 AlphaXiv
PRIMT: Preference-based Reinforcement Learning with Multimodal Feedback and Trajectory Synthesis from Foundation Models figure
AlphaXiv 中文论文页面（可滚动查看）