SOURCE-LINKED INTELLIGENCE

Real-World Reinforcement Learning with MPC Scaffolding for Dexterous Manipulation

arXiv · AI, language, vision and robotics · article · Sep 14, 2026 · UTC

Real-world reinforcement learning (RL) offers a promising route to dexterous manipulation policies that can adapt directly from physical interaction, but learning is hindered by inefficient early exploration and costly failures. We propose a framework that uses sampling-based model predictive control (MPC) as scaffolding for real-world dexterous RL, providing structured prior experience and task-directed guidance during learning without human demonstrations or corrective actions. A small set of MPC trajectories is first used to populate an offline replay buffer and to pretrain the actor and cr

Read original source ↗ Open in workspace

recordType

paper

region

Global