Rethinking World-Action Model for Compositional and In-Context Robotic Manipulation

Shukai Gong, Xuanran Zhai, Yintianrun Zhang, Ruopeng Cui, Ye Huang, Yiyang Fu, Dexuan Lyu, Chaojie Li, Xinyi Song, Peiwen Lin, Chuang Wang, Mingyuan Jia, Yufan Deng, Jiaxin Fang, Bo Liang, Jiaxin Li, Yuxiang Gao, Hao Liu, Daquan Zhou · N/A · 2026

Summary

Long-horizon compositional manipulation has become increasingly important for real-world robot deployment, where a single task involves multiple coordinated subtasks.

Abstract Summary

Long-horizon compositional manipulation has become increasingly important for real-world robot deployment, where a single task involves multiple coordinated subtasks. Existing world-action models (WAMs) jointly predict short-horizon visual futures and actions, but typically lack explicit subtask-level reasoning. We propose Visual Goal-conditioned Action Reasoning (ViGAR), a hierarchical framework that factorizes manipulation into a visual subgoal planner and a subgoal executor. Given the current observation and global instruction, the subgoal planner predicts a visual subgoal for the next subtask. The subgoal executor then jointly generates future visual trajectories and actions conditioned on the predicted subgoal. Both components share a pretrained world-model representation, enabling task-level planning and action generation to benefit from common physical knowledge. Moreover, our framework naturally supports in-context learning: using a global goal image as context can induce different subtask decompositions and behaviors without parameter updates. On the RoboTwin Clean2Random benchmark, ViGAR achieves 82.00% and 67.02% success rates under the Clean and Random settings, respectively, surpassing the strongest baseline by 12.86 percentage points in average success rate. Real-world robot experiments on five compositional and two in-context learning tasks further confirm the effectiveness of ViGAR.

Key Points

Long-horizon compositional manipulation has become increasingly important for real-world robot deployment, where a single task involves multiple coordinated subtasks.
Existing world-action models (WAMs) jointly predict short-horizon visual futures and actions, but typically lack explicit subtask-level reasoning.
We propose Visual Goal-conditioned Action Reasoning (ViGAR), a hierarchical framework that factorizes manipulation into a visual subgoal planner and a subgoal executor.
Given the current observation and global instruction, the subgoal planner predicts a visual subgoal for the next subtask.
The subgoal executor then jointly generates future visual trajectories and actions conditioned on the predicted subgoal.

Preliminaries

Consider a robot task specified by a language instruction $\ell$.

At inference step $t$, a VLA policy $\pi^{\text{VLA}}_{\theta}$ with parameters $\theta$ predicts an $H_{a}$-step action chunk $\mathbf{A}_{t+1}:=\mathbf{a}_{t+1:t+H_{a}}\in\mathbb{R}^{H_{a}\times d_{a}}$ conditioned on the current multi-view observation $\mathbf{o}_{t}=\{\mathbf{o}_{t}^{v}\}_{v=1}^{V}$, proprioceptive state $\mathbf{s}_{t}\in\mathbb{R}^{d_{s}}$, and textual metadata $\mu$ specifying the embodiment, camera resolution, and control mode: where $H_{a}$ denotes the a...

Preliminaries — Rethinking World-Action Model for Compositional and In-Context Robotic Manipulation — Figure 2 : Model architecture of ViGAR: (1) the subgoal planner $\mathcal{G}_{\theta}$ generates the next subgoal image from the task instruction and current observation, optionally conditioned on a global goal image for in-context learning (ICL); (2) the policy $\pi^{\text{WAM}}_{\phi}$ takes in the predicted subgoal and jointly predicts the future frames and the action chunk.

Method

In this section, we describe the two essential parts of ViGAR: a subgoal planner and a subgoal-guided world-action model. We divide each compositional manipulation task into multiple subtasks, where each subtask contains an atom action and a subgoal $(\ell_{i},\mathbf{g}_{i})$. Our subgoal planner generates a subgoal for each subtask, as illustrated in Section 4.1 .

Then, our world-action policy predicts the action chunk from the current observation, proprioceptive state and generated subgoal, as detailed in Section 4.2 .

Method — Rethinking World-Action Model for Compositional and In-Context Robotic Manipulation — (a) Subgoal look-ahead rule ( Equation 2 ): redirecting the last $p\%$ of $T_{i}$ to $\mathbf{g}_{i+1}$ makes the supervision consistent across the boundary.

Experiments

Implementation. The subgoal planner $\mathcal{G}_{\theta}$ and subgoal-guided world-action model $\pi^{\text{WAM}}_{\phi}$ of ViGAR are initialized from the pretrained Cosmos3-Nano checkpoint [ 1 ] , a 16B-A8B-parameter omnimodal world model that adopts a dual-branch mixture-of-transformers architecture.

The subgoal planner $\mathcal{G}_{\theta}$ is adapted from the pretrained backbone by transforming its image-to-video generation capability into image editing.

Experiments — Rethinking World-Action Model for Compositional and In-Context Robotic Manipulation — (b) End-effector ROI weighting ( Equation 3 ): the planning loss is weighted toward the end-effector region-of-interest (ROI), so that supervision concentrates where manipulation actually happens rather than on static background.

Sources and Demonstrations

Figures are reproduced from Gong et al. See the full paper for experimental details and the project page, when the paper links one, for demonstrations and videos.