analiza aiqbench
The paper proposes a robot learning pipeline that uses a small amount of human play to train a behavior-cloned play policy, which then collects interaction data for a learned world model and guides model-predictive control. Its main claim is data efficiency: the authors report comparable success to reinforcement learning with 490 times less interaction data, though the evidence comes from specific tasks (underactuated pushing and in-hand reorientation) and a single multifingered hand. For readers, the practical question is whether world-model-based planning can reduce the real-robot data burden without task-specific demonstrations, and that depends on how well the results transfer beyond these benchmarks.
Dexterity from Supervised Play
arXiv:2610.22340v1 Announce Type: cross Abstract: How can robots autonomously collect the contact-rich experience needed to learn dexterous manipulation? We present DSP: Dexterity from supervised play, a framework in which a robot learns a play policy from minutes of human play through behavior cloning, then deploys it autonomously to collect hours of interaction data to train a world model. During inference, the same play policy then guides model-predictive control (MPC) by proposing candidate action sequences for evaluation under the learned world model. To accelerate inference, we also introduce a method for distilling a closed-loop, goal-conditioned policy through self-play within the learned world model. We evaluate this framework on underactuated pushing and any-to-any in-hand pose reorientation with a multifingered hand, in simulation and real world. In real world experiment, our method outperforms hindsight behavior cloning by 69.6% in success rate and achieves comparable success to reinforcement learning with 490 times less interaction data. These results demonstrate that supervised play can autonomously collect interaction data, guide model-based planning, and enable policy learning without any task-specific demonstrations or sim-to-real transfer.
Abstract:How can robots autonomously collect the contact-rich experience needed to learn dexterous manipulation? We present DSP: Dexterity from supervised play, a framework in which a robot learns a play policy from minutes of human play through behavior cloning, then deploys it autonomously to collect hours of interaction data to train a world model. During inference, the same play policy then guides model-predictive control (MPC) by proposing candidate action sequences for evaluation under the learned world model. To accelerate inference, we also introduce a method for distilling a closed-loop, goal-conditioned policy through self-play within the learned world model. We evaluate this framework on underactuated pushing and any-to-any in-hand pose reorientation with a multifingered hand, in simulation and real world. In real world experiment, our method outperforms hindsight behavior cloning by 69.6% in success rate and achieves comparable success to reinforcement learning with 490 times less interaction data. These results demonstrate that supervised play can autonomously collect interaction data, guide model-based planning, and enable policy learning without any task-specific demonstrations or sim-to-real transfer.
Robotics (cs.RO); Artificial Intelligence (cs.AI)
arXiv:2610.22340 [cs.RO]
(or arXiv:2610.22340v1 [cs.RO] for this version)
https://doi.org/10.48550/arXiv.2610.22340 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 · 267 cuvinte
Subiecte conexe
The Machines that Make the Machines
NVIDIA · 7 oct. 2026
Spark-0: Scaling Spatially Aligned Tactile Pretraining for Robot Dexterous Manipulation
arXiv · 16 oct. 2026
Synthesis Superintelligence: from Semiconductors to Superconductors — Periodic Labs’ Liam Fedus and Ekin Dogus Cubuk
Latent Space · 8 oct. 2026
Import AI 463: Self-improving robots; a 10k Chinese GPU cluster; and an elegiac essay for the human era
Import AI · 29 iun. 2026