We trained a neural network to play Minecraft by Video PreTraining (VPT) on a massive unlabeled video dataset of human Minecraft play, while using only a small amount of labeled contractor data. With fine-tuning, our model can learn to craft diamond tools, a task that usually takes proficient humans over 20 minutes (24,000 actions).
Despite the recent promise in robot control, video generative models suffer from a domain mismatch due to their primary focus on content creation. For example, their design inherently prioritizes visual fidelity and creativity over computational efficiency and physical realism.
arXiv:2606. 00054v1 Announce Type: cross Abstract: Recent progress in generalizable embodied control has been driven by large-scale pretraining of Vision-Language-Action (VLA) models.
arXiv:2609.09059v1 Announce Type: new Abstract: While many video-game environments (VGEs) have played crucial roles in advancing reinforcement learning (RL), developing novel VGEs or modifying existi...
arXiv:2607. 06988v1 Announce Type: cross Abstract: Steering robot foundation models (RFMs) toward new task variants or user-preferred behaviors remains challenging, often requiring additional robot demonstrations, task-specific fine-tuning, or long-context conditioning.
arXiv:2609.01560v1 Announce Type: cross Abstract: We present H3-World, an efficient framework that turns the 33B MiniMax-H3 video generator into an interactive world model. Our key finding is that, a...
arXiv:2609.37250v1 Announce Type: cross Abstract: World-action models (WAMs) couple future visual-state prediction with action generation. By adapting video generators or image-editing models pretrai...
arXiv:2608. 14022v1 Announce Type: cross Abstract: Action-conditioned video world models require low-latency causal generation and reliable responses to game-native controls.
We present LingBot-World 2. 0 (also known as LingBot-World-Infinity), an advanced iteration of LingBot-World featuring four distinct upgrades.
Zing-0.5 is a 5B autoregressive world model that enables users to explore and influence generated worlds through joint keyboard and online text control. It integrates unified action and text conditioning, event-scale supervision for incremental generation, and low-cost real-time interaction, achieving high scores on WBench Navigation. The authors release model weights, inference code, and a serving implementation to support further research on playable generated worlds.
Generalizable robot manipulation requires policies that can anticipate how visual scenes evolve while executing language instructions. While recent Vision-Language-Action models benefit from large-scale pretraining, their predominantly static pretraining objectives provide limited supervision for physical dynamics and temporal causality, leaving control-relevant knowledge to be learned from downstream robot demonstrations.
SimWAM is a lightweight World-Action Model that uses future‑video prediction only during training to supervise an action expert, enabling end‑to‑end autonomous driving without costly test‑time future imagination. The architecture co‑trains a pretrained video expert and a lightweight action expert via joint flow matching, while an isolated attention mask keeps action prediction independent of future frames. This design allows the video backbone to be swapped and the action expert scaled independently, achieving 91.5 PDMS on NAVSIM, outperforming state‑of‑the‑art WAM planners with lower latency and zero‑shot transfer to nuScenes.
ZimaBlue is a scalable framework that learns generalizable World Action Models (WAMs) from large-scale egocentric videos. It follows a three-stage curriculum: causal video pre‑training, video‑action mid‑training with a unified action representation, and final specialization to a target robot. The system employs an asynchronous Slow‑Fast architecture to enable real‑time 30 Hz action prediction, achieving a jump in real‑robot zero‑shot success from 36.1% to 77.8% when leveraging over 120,000 hours of embodied video.