analiza aiqbench
VLaRL addresses the critical sim-to-real gap in robotics by leveraging frozen vision-language-action models, allowing residual reinforcement learning policies to be trained safely in simulation and transferred via latent space alignment rather than pixel-level matching. This approach significantly improves success rates on contact-rich manipulation tasks without requiring costly real-world RL training or online adaptation, offering a practical pathway for deploying robust robotic agents.
VLaRL: Augmenting Vision-Language-Action Models with Simulation-Trained Latent-Conditioned Residual RL
arXiv:2609.30868v1 Announce Type: new Abstract: Vision-language-action (VLA) models provide broad, instruction-conditioned manipulation behaviors, but their physical execution can remain imprecise during contact-rich interaction. Residual reinforcement learning (RL) can correct such errors while keeping the VLA frozen, but real-robot RL is costly and safety-critical. We propose VLA Latent-Conditioned RL (VLaRL), which enables residual RL for frozen VLAs to be trained in simulation and deployed on real robots without real-world RL or online adaptation. The key challenge is transferring the learned residual policy despite the visual gap between simulation and reality. Rather than requiring pixel-level visual correspondence, VLaRL uses the VLA's internal vision-language latent representation to condition residual control and as the sim-to-real transfer interface, and learns a lightweight mapper that transforms simulation-derived latents toward the real latent distribution. Across four contact-rich manipulation tasks and two VLA backbones, VLaRL improves real-world success in all task-backbone combinations, while controlled ablations demonstrate the importance of both latent conditioning and latent alignment for transferring simulation-trained residual control.
Abstract:Vision-language-action (VLA) models provide broad, instruction-conditioned manipulation behaviors, but their physical execution can remain imprecise during contact-rich interaction. Residual reinforcement learning (RL) can correct such errors while keeping the VLA frozen, but real-robot RL is costly and safety-critical. We propose VLA Latent-Conditioned RL (VLaRL), which enables residual RL for frozen VLAs to be trained in simulation and deployed on real robots without real-world RL or online adaptation. The key challenge is transferring the learned residual policy despite the visual gap between simulation and reality. Rather than requiring pixel-level visual correspondence, VLaRL uses the VLA's internal vision-language latent representation to condition residual control and as the sim-to-real transfer interface, and learns a lightweight mapper that transforms simulation-derived latents toward the real latent distribution. Across four contact-rich manipulation tasks and two VLA backbones, VLaRL improves real-world success in all task-backbone combinations, while controlled ablations demonstrate the importance of both latent conditioning and latent alignment for transferring simulation-trained residual control.
Robotics (cs.RO)
arXiv:2609.30868 [cs.RO]
(or arXiv:2609.30868v1 [cs.RO] for this version)
https://doi.org/10.48550/arXiv.2609.30868 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 250 cuvinte
Subiecte conexe
Physical AI’s moment has arrived – but moving from demo to deployment is the hard part. AWS wants to fix that
ZK Research · 24 aug. 2026
Build a Physical AI model factory with NVIDIA Cosmos 3 on SageMaker HyperPod
AWS · 4 sept. 2026
NVIDIA Enables the Next Era Of Physical AI Research With Agent Skills For Autonomous Vehicles, Robotics And Vision AI
NVIDIA · 3 iun. 2026
From Pixels to Executable Physics: Behavior-Grounded World Models for Embodied Action Consequence Prediction
arXiv · 14 oct. 2026