It is shown that JEPA features outperform video VAE latents, most clearly out of distribution, and that training-time future prediction choices significantly impact the policy even when future prediction is skipped at inference, suggesting that robot foundation models may benefit from architectures designed for control...
Mateo Clémente, George Z. Xue, Amin Abyaneh et al.· 0 citations
Spatial Grafting is proposed, a versatile, lightweight spatial module that binds frozen reconstruction features to metric, robot-relative geometry and injects them into the flow-matching action expert through cross-attention without modifying the host's perceptual pathway, so the host retains the full benefit of its pr...
A novel vision-language model (VLM)-based navigation framework that addresses challenges by adaptively retrieving task-relevant cues from a global memory module and integrating them with the agent's egocentric observations and enhances spatial reasoning and decision-making in long-horizon tasks.
Faster-WAM, an instantiation of DoT for WAMs, which docks a single-layer action head onto a 30-layer video backbone, achieves competitive performance on LIBERO and RoboTwin 2.0 while demonstrating strong out-of-distribution generalization on LIBERO-Plus.
Liheng Ma, Rui Yang, Zhan-Guang Zhang et al.· 4 citations
KAM-WM is presented, a framework that extracts a coarse directional interaction cue from a frozen latent video world model without rollout or world-model fine-tuning and indicates that a frozen video model can provide a useful first-order visual prior for control without the test-time cost of future rollout.
RoboHarness is proposed, a unified framework that encapsulates independently developed robot control systems as reusable agentic skills as well as a general framework compatible with a broader range of robot policies, such as navigation policies, model predictive controllers, and world-action models.
Jinbang Huang, Yuan Hu, Zhiyuan Li et al.· arXiv.org· 3 citations· ⚡1
Future-State-Conditioned VLN (FSC-VLN), a deployable model that augments a causal policy with a future-query token and uses training-only future-state supervision to distill information from future observations into the policy state, is proposed.
Lingfeng Zhang, Zhanguang Zhang, Liheng Ma et al.· arXiv.org· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.