WoVR+: World Model as Reliable Simulator for Versatile VLA Policy Optimization with RL

Shangqing Zhou; Zhennan Jiang; Ning Ma; Xin Fu; Yupeng Zheng; Shuai Tian; Huangrui Li; Jiahao Liu; Chao Yu; Zhongpu Xia; Haoran Li; Dongbin Zhao

arXiv · 2610.17922

WoVR+: World Model as Reliable Simulator for Versatile VLA Policy Optimization with RL

Shangqing Zhou·Zhennan Jiang·Ning Ma·Xin Fu·Yupeng Zheng·Shuai Tian·Huangrui Li·Jiahao Liu·Chao Yu·Zhongpu Xia·Haoran Li·Dongbin Zhao

View original source ↗Read source document ↗Export metadata ↓

Abstract

Action-conditioned world models can reduce costly real-robot interaction in reinforcement learning for Vision-Language-Action models (VLA-RL), but hallucinations can mislead policy optimization. Extending them to richer VLA interfaces and long-horizon tasks raises challenges in spatiotemporal and cross-modal consistency, exploration-reliability balance in RL, and compounding rollout errors. We present WoVR+, extending WoVR's hallucination-aware WM-RL framework by redesigning the learned environment and imagined interaction protocol: A VLA-interface-aligned world model provides multi-view prediction and proprioceptive updates while mitigating hallucinations. Chunk-wise noise preserves flow-based policy exploration while limiting prediction-degrading action jitter. For long-horizon tasks, stage-local imagined training combines stage-entry resets, stage-specific rewards, local horizons, and controllable sampling to shorten generated prefixes and balance training across stages. WoVR+ improves success over SFT by 21.09% and 16.7% on RoboTwin and 6 real-robot tasks. Long-horizon phone-packaging success rises from 30% to 75%. Videos are available at https://youmo445.github.io/WoVR_Plus.github.io/.

Explore related subjects

cs.RO

Keep this discovery

Explore connections, maps & timelines

BibTeX RIS

Shangqing Zhou, Zhennan Jiang, Ning Ma, Xin Fu, Yupeng Zheng, Shuai Tian, Huangrui Li, Jiahao Liu, Chao Yu, Zhongpu Xia, Haoran Li, Dongbin Zhao. 2026-10-13. WoVR+: World Model as Reliable Simulator for Versatile VLA Policy Optimization with RL. https://arxiv.org/abs/2610.17922

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

All-Electric Heavy-Duty Robotic Manipulator: Surrogate-Enhanced Modeling and Adaptive Modular Control

This paper presents a unified system-level modeling and control framework for an all-electric heavy-duty robotic manipulator (HDRM) driven by electromechanical linear actuators (EMLAs). A surrogate-enhanced actuator model, combining integrated electromechanical dynamics with a neural network trained on a dedicated testbed, is integrated into an extended virtual decomposition control (VDC) architecture augmented by a natural adaptation law. The derived analytical HDRM model supports a hierarchical control structure that seamlessly maps high-level force and velocity objectives to real-time actuator commands, accompanied by a Lyapunov-based stability proof. In multi-domain simulations of both cubic and a custom planar triangular trajectory, the proposed adaptive modular controller achieves sub-centimeter Cartesian tracking accuracy. Experimental validation of the same 1-DoF platform under realistic load emulation confirms the efficacy of the proposed control strategy. These findings demonstrate that a surrogate-enhanced EMLA model embedded in the VDC approach can enable modular, real-time control of an all-electric HDRM, supporting its deployment in next-generation mobile working machines.

cs.RO↗

arXiv2026

METIS: Multi-Source Egocentric Training for Integrated Dexterous Vision-Language-Action Model

Building a generalist robot that can perceive, reason, and act across diverse tasks remains an open challenge, especially for dexterous manipulation. A major bottleneck lies in the scarcity of large-scale, action-annotated data for dexterous skills, as teleoperation is difficult and costly. Human data, with its vast scale and diverse manipulation behaviors, provides rich priors for learning robotic actions. While prior works have explored leveraging human demonstrations, they are often constrained by limited scenarios and a large visual gap between human and robots. To eliminate these limitations, we propose METIS, a vision-language-action (VLA) model for dexterous manipulation pretrained on multi-source egocentric datasets. We first construct EgoAtlas, which integrates large-scale human and robotic data from multiple sources, all unified under a consistent action space. We further extract motion-aware dynamics, a compact and discretized motion representation, which provides efficient and expressive supervision for VLA training. Built upon them, METIS integrates reasoning and acting into a unified framework, enabling effective deployment to downstream dexterous manipulation tasks. Our method demonstrates exceptional dexterous manipulation capabilities, achieving highest average success rate in six real-world tasks. Experimental results also highlight the superior generalization and robustness to out-of-distribution scenarios. These findings emphasize METIS as a promising step toward a generalist model for dexterous manipulation.

cs.RO↗

arXiv2026

Hiking in the Wild: A Scalable Perceptive Parkour Framework for Humanoids

Achieving robust humanoid locomotion in complex, unstructured environments requires transitioning from reactive proprioception to proactive perception. However, integrating exteroception remains a significant challenge: existing approaches often fail to achieve high-speed, stable locomotion across diverse, unstructured terrains. In this work, we present Hiking in the Wild, an extensible, end-to-end perceptive legged parkour framework designed for robust humanoid hiking. Our approach maps depth inputs and proprioception directly to joint actions, without relying on external localization and mapping. To ensure safety and training stability, we introduce two key mechanisms: a foothold safety mechanism combining extensible Terrain Edge Detection with Foot Volume Points to reduce catastrophic slippage on edges, and a Flat Patch Sampling strategy that mitigates reward hacking by generating feasible navigation targets. Extensive experiments on the Unitree G1 humanoid robot demonstrate that our policy enables robust traversal of complex terrains at speeds up to 2.5 m/s.

cs.RO↗