Play2Perfect: What Matters in Dexterous Play Pretraining for Precise Assembly?

Hugging Face Daily Papers Papers
reinforcement-learning robotics assembly dexterous-manipulation pretraining sim-to-real sample-efficiency

Summary

Play2Perfect is a reinforcement learning framework that uses playful interaction with diverse objects to learn general manipulation skills, then fine-tunes for precise assembly tasks, achieving 33x sample efficiency and zero-shot sim-to-real transfer on tight insertions.

Multi-fingered robots promise the speed and dexterity of human hands, yet challenging problems such as precise assembly have remained out of reach. These tasks are contact-rich, making data collection for imitation learning difficult, and sparse-reward, making direct exploration with reinforcement learning (RL) intractable. Consequently, prior work has made progress by structuring the problem with specialized grippers, tool attachments, and environment fixtures. In this work, we argue that before a robot can perfect precise assembly, it must first learn to play. We further ask the question: what factors in the process of learning to play matter for precise assembly? We propose Play2Perfect, an RL framework for task-agnostic pretraining through play on diverse objects and goals, which is then perfected on precise assembly. The goal of play is to acquire reusable manipulation priors, such as grasping, in-hand reorientation and pose reaching. Finetuning then adapts this general prior to assembly, focusing exploration on the final contact-rich, high-precision interactions needed for success. We systematically study key design choices in play pretraining, including object diversity, training objective, trajectory diversity, and goal precision. We show that our prior is 33x more sample-efficient than RL training from scratch, even when provided with dense, multi-stage rewards. We demonstrate zero-shot sim-to-real transfer, achieving 60% success on tight insertions with only 0.5 mm contact clearance, and over 50% success on long-horizon multi-part assembly and screwing.
Original Article
View Cached Full Text

Cached at: 07/01/26, 07:44 PM

Paper page - Play2Perfect: What Matters in Dexterous Play Pretraining for Precise Assembly?

Source: https://huggingface.co/papers/2606.26428

Abstract

A reinforcement learning framework called Play2Perfect enables sample-efficient robotic assembly tasks by first learning general manipulation skills through playful interaction with diverse objects, then adapting these skills for precise assembly through fine-tuning.

Multi-fingered robots promise the speed and dexterity of human hands, yet challenging problems such as preciseassemblyhave remained out of reach. These tasks are contact-rich, making data collection forimitation learningdifficult, andsparse-reward, making direct exploration withreinforcement learning(RL) intractable. Consequently, prior work has made progress by structuring the problem with specialized grippers, tool attachments, and environment fixtures. In this work, we argue that before a robot can perfect preciseassembly, it must first learn to play. We further ask the question: what factors in the process of learning to play matter for preciseassembly? We propose Play2Perfect, an RL framework fortask-agnostic pretrainingthrough play on diverse objects and goals, which is then perfected on preciseassembly. The goal of play is to acquire reusablemanipulation priors, such asgrasping,in-hand reorientationandpose reaching. Finetuning then adapts this general prior toassembly, focusing exploration on the final contact-rich, high-precision interactions needed for success. We systematically study key design choices in play pretraining, including object diversity, training objective, trajectory diversity, and goal precision. We show that our prior is 33x more sample-efficient than RL training from scratch, even when provided with dense, multi-stage rewards. We demonstrate zero-shotsim-to-real transfer, achieving 60% success on tight insertions with only 0.5 mm contact clearance, and over 50% success on long-horizon multi-partassemblyand screwing.

View arXiv pageView PDFProject pageGitHub13Add to collection

Get this paper in your agent:

hf papers read 2606\.26428

Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash

Models citing this paper0

No model linking this paper

Cite arxiv.org/abs/2606.26428 in a model README.md to link it from this page.

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2606.26428 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2606.26428 in a Space README.md to link it from this page.

Collections including this paper0

No Collection including this paper

Add this paper to acollectionto link it from this page.

Similar Articles

Learning dexterity

OpenAI Blog

OpenAI announces Dactyl, a system that learns robotic hand dexterity through simulation and reinforcement learning, using LSTMs to generalize across different physical environments and the Rapid PPO implementation to train policies that transfer to real-world manipulation tasks.

Zero-Shot Sim-to-Real Robot Learning: A Dexterous Manipulation Study on Reactive Catching

Hugging Face Daily Papers

This paper introduces Domain-Randomized Instance Set (DRIS), a method that simultaneously represents multiple randomized instances to improve sim-to-real transfer for dexterous manipulation. It demonstrates zero-shot transfer on a challenging reactive catching task with a flat plate end-effector, requiring no real-world fine-tuning.

Sample-Efficient Post-Training for LEGO Spatial-Physics Reasoning

arXiv cs.LG

This paper identifies a failure mode called PhysHack in LLM-based LEGO assembly generation and proposes PVPO, a sample-efficient reinforcement learning method with model-based data selection that improves physical and semantic alignment using only a small fraction of training data.