Skip to content

Copper-Policy: Focus on the Representation for Robust Robot Manipulation

Sep 2026 · arXiv.org · Vol abs/2609.32779 · 1 citation · 51 references
Computer Science

TL;DR

Copper-Policy is introduced, which learns a compact World representation with the policy rather than relying on a predefined target space and predicts future observation embeddings conditioned on task intention without reconstructing pixels through temporal joint-embedding prediction.

Abstract

World Action Models (WAMs) acquire behavioral priors by modeling future scene evolution, but predicting detailed futures in pixel or latent space incurs substantial cost. Recent evidence that co-training gains persist without test-time generation raises a question: what must a WAM learn to improve control? We introduce Copper-Policy, which learns a compact World representation with the policy rather than relying on a predefined target space. Through temporal joint-embedding prediction, it predicts future observation embeddings conditioned on task intention without reconstructing pixels. This prediction and action decoding shape the representation jointly, while the policy retains access to current-frame spatial detail for execution. Representation analyses show that the learned features better separate task-driven change from perturbations and provide complementary information for control. Compact prediction targets reduce training tokens per sample, enabling a 2B-parameter model trained in 9.67 hours on 8$\times$ RTX 5090 GPUs and 6$\times$ faster than Fast-WAM on matched A100 GPUs. Copper-Policy outperforms every compared method without embodied pretraining on RoboTwin and several embodied-pretrained VLAs on LIBERO-Plus (80.85%). On three challenging real-robot tasks, it performs comparably to $\pi_{0.5}$ and attains a higher average score. Together, these results show that Copper-Policy combines strong control performance with efficient training.

View source

Similar papers

Sep 2026

Rethinking Representations for World-Action Modeling

World-action models jointly learn robot policies and predict future observations, making the representation space an interface between control and prediction. We study the design of this space through controlled comparisons, finding that neither reconstruction fidelity nor pre-trained perceptual features alone ensure e...

Hao-Yi Jiang, Liu Liu, Xin-Jiang Wang et al. · 5 citations
Preprint Oct 2026

MiniWAM: Learning Compact Future Targets for Efficient World-Action Modeling

It is demonstrated that effective world-action modeling does not require predicting native visual futures, and that compact predictive representations provide a strong and substantially more efficient target for policy learning.

Jie Chen, Ruofei Bai, Yu-Xin Cai et al. · 0 citations
Preprint Aug 2026

PredVLA: Predictive Sensorimotor Modeling for Sub-Million-Parameter Robot Manipulation

A mechanism-by-mechanism transition to the recurrent behavior-cloning baseline shows that replacing the predictive pathway with direct observation input produces the largest single performance drop, accounting for approximately $70\% of the endpoint gap.

Hiroki Sawada, Shunichi Kasahara · 1 citation
Sep 2026

WorldLine: Action-Driven Visual Simulation for Robotic Manipulation

WorldLine is introduced, an action-driven visual simulator that decouples transferable dynamics learning from heterogeneous action grounding and makes WorldLine a scalable and efficient visual simulator for policy evaluation and embodied planning.

Shenghe Zheng, Wen-Bo Li, Ji-Yao Zhang et al. · 1 citation
#artificial intelligence Preprint Aug 2026

PAVE: Predictive Alignment and Value-Guided Evolution for World-Action Policies

Every valid trajectory can teach what physically happened, while the actor is deployed only under the condition associated with relatively better actions, in \method, a direct world-action policy that combines outcome-agnostic predictive learning with outcome-aware policy improvement.

Bo-Tong Zhao, Fangjie Yu, Tim Yu et al. · 0 citations
Preprint Oct 2026

IronMan: Information-Constrained Video-Action Learning for Robot Manipulation

Video Action Models (VAMs) couple visual dynamics modeling with action generation for robot manipulation. However, video representations are not naturally suited to action generation, as exposing the action policy to excessive visual detail can impair its generalization ability. Therefore, we introduce IronMan (Informa...

Yuan-Shuo Zhang, Wen-Zhe Zhao, Zi-Xing Lei et al. · 0 citations

Related blog posts

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.