LingBot-VA: Causal World Modeling for Robot Control

作者：alongside vision-language pre-training, establishes a fresh, independent foundation for robot learning. Intuitively, visual dynamics. Inspired by this, we introduce LingBot-VA, integrating vision, action tokens, driven by a Mixture-of-Transformers (MoT) architecture, a closed-loop rollout mechanism, an asynchronous inference pipeline, parallelizing action prediction, real-world scenarios, data efficiency in post-training, strong generalizability to · 会议/期刊：arXiv · 日期：2026-01-29 · 来源：Low-Level Learning-Based Action Modelling / Input Modelling / 2D Vision Language Action Models with Auxiliary Tasks - World Model & Visual Prediction / World Model