Dexterity from Supervised Play

Botao He, Yiming Dou, Yiannis Aloimonos, Jitendra Malik, Akash Sharma, Ruoshi Liu · N/A · 2026

Summary

How can robots autonomously collect the contact-rich experience needed to learn dexterous manipulation?

Abstract Summary

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.

Key Points

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.

Method

Given a dexterous manipulation task, our pipeline \dspfont DSP achieves it in 4 stages described in Fig. 2 . We first collect data through teleoperation to train a play policy (see Sec. III-B ). We then rollout this policy on the robot autonomously to collect a play dataset that contains observation–action transitions (see Sec. III-C ).

This play dataset is used to train an action-conditioned world model to model the dynamics of the interaction (see Sec. III-D ).

Method — Dexterity from Supervised Play — Fig. 1: \dspfont DSP Pipeline Overview. (1) Human teleoperation trains a goal-agnostic play policy. (2) The policy collects interaction data through autonomous play. (3) These data train a world model to predict future states from past observations and proposed actions. (4) MPC uses the model to select actions for a requested goal. (5) Optionally, we can distill a goal-conditioned policy from self-play in the learned world model.

Experiments

Our experiments evaluate whether \dspfont DSP can turn goal-free play into goal-directed manipulation across tasks and on real hardware, whether play-guided MPC is more interaction-efficient than direct policy learning and task-specific RL, and how play-policy design, autonomous-play scale, and world-model design affect prediction and control. We organize the evaluation around three tasks.

Task A: Push-T tests state-based planar pushing to target poses under a fixed tight 10k-frame teleoperation budget, measuring success and best mask overlap.

Experiments — Dexterity from Supervised Play — Fig. 2: Dexterity from Supervised Play. (a) A goal-agnostic play policy is trained by behavior cloning on human teleoperation data. (b) The learned policy interacts autonomously with the environment to collect observation–action transitions. (c) These transitions train an action-conditioned world model to predict future observations. (d) At inference time, the play policy proposes candidate actions, the world model predicts their outcomes, and model-predictive control selects the candidate with the highest goal-conditioned reward. The selected action is executed in the environment, and planning repeats from the new observation.

Sources and Demonstrations

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