ADEPT: Accelerating Dexterity via Pre-Training and Post-Training using Reinforcement Learning

• URL: http://arxiv.org/abs/2608.19182v1
• Date: Wed, 19 Aug 2026 17:55:06 GMT
• Title: ADEPT: Accelerating Dexterity via Pre-Training and Post-Training using Reinforcement Learning
• Abstract summary: We introduce Accelerating Dexterity via Pre-Training (ADEPT), a large-scale reinforcement learning framework for learning sim-to-real transferable dexterity.<n>ADEPT pretrains a dexterous policy on a generic object reposing task, then post-trains downstream policies with this pretrained behavior as a prior.<n>We distill post-trained teachers into perceptive students that zero-shot sim-to-real transfer on two embodiments.
• Score: 4.555414332509501
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: We introduce Accelerating Dexterity via Pre-Training (ADEPT), a large-scale reinforcement learning (RL) framework for learning sim-to-real transferable dexterity across high degree-of-freedom (DoF) robot embodiments that can solve long-horizon tasks directly from raw visuo-tactile perception. ADEPT pretrains a dexterous policy on a generic object reposing task, then post-trains downstream policies with this pretrained behavior as a prior. ADEPT enables learning new behaviors that are otherwise difficult to discover from scratch on multi-fingered robots and avoids learning the same set of skills over again for every new downstream task. The pretrained policy zero-shots the reposing phase of downstream tasks, but naïve RL fine-tuning rapidly degrades this capability during transfer. We address this with a stable post-training recipe combining behavior-cloning distillation, critic warm-up, and conservative on-policy updates. To safely exploit the full kinematic dexterity, we introduce a joint-space Geometric Fabric that mediates between the RL policy and the robot. We distill post-trained teachers into perceptive students that zero-shot sim-to-real transfer on two embodiments: a 23 DoF Kuka-Allegro with two RGB cameras, and a 29 DoF Flexiv-Sharpa with two RGB cameras and five vision-based tactile sensors, and can solve long-horizon tasks from challenging initial states with dexterity at human-level speed.

Related papers

• Pre-training Visual Dexterity in Simulation [67.17477115017896]
  We introduce a pre-training framework for dexterous manipulation that uses data entirely collected in simulation.<n>With the help of five operators, we collect 75 hours of multi-task dexterous manipulation over one week, and use it to pre-train a causal transformer on a sequence modeling objective.<n>We find that our approach outperforms training behavior cloning policies from scratch, showing that simulation teleoperation is a viable pre-training source for real-world dexterous manipulation.
  arXiv  Detail & Related papers  (2026-08-16T20:21:33Z)
• FlashSAC: Fast and Stable Off-Policy Reinforcement Learning for High-Dimensional Robot Control [55.38832429564216]
  Reinforcement learning (RL) is a core approach for robot control when expert demonstrations are unavailable.<n>We present FlashSAC, a fast and stable off-policy RL algorithm built on Soft Actor-Critic.<n>Across over 60 tasks in 10 simulators, FlashSAC consistently outperforms PPO and strong off-policy baselines in both final performance and training efficiency.
  arXiv  Detail & Related papers  (2026-04-06T09:03:41Z)
• PLATE: Plasticity-Tunable Efficient Adapters for Geometry-Aware Continual Learning [1.8985807299725523]
  We develop a continual learning method for pretrained models that requires no access to old-task data.<n> redundancy offers a natural bias for emphwhere to place plasticity.<n>textscPLATE provides explicit control over the plasticity-retention trade-off.
  arXiv  Detail & Related papers  (2026-02-03T18:59:42Z)
• Opening the Sim-to-Real Door for Humanoid Pixel-to-Action Policy Transfer [59.02729900344616]
  GPU-accelerated, photorealistic simulation has opened a scalable data-generation path for robot learning.<n>We develop a teacher-student-bootstrap learning framework for vision-based humanoid loco-manipulation.<n>This represents the first humanoid sim-to-real policy capable of diverse articulated loco-manipulation using pure RGB perception.
  arXiv  Detail & Related papers  (2025-11-30T20:07:13Z)
• Learned Controllers for Agile Quadrotors in Pursuit-Evasion Games [42.74003740156243]
  We address the problem of agile 1v1 quadrotor pursuit-evasion, where a pursuer and an evader learn to outmaneuver each other.<n>We propose an Asynchronous Multi-Stage Population-Based (AMSPB) algorithm to tackle these issues.<n>Within this framework, we train neural network controllers that output either velocity commands or body rates with collective thrust.
  arXiv  Detail & Related papers  (2025-06-03T13:19:23Z)
• DexterityGen: Foundation Controller for Unprecedented Dexterity [67.15251368211361]
  Teaching robots dexterous manipulation skills, such as tool use, presents a significant challenge.<n>Current approaches can be broadly categorized into two strategies: human teleoperation (for imitation learning) and sim-to-real reinforcement learning.<n>We introduce DexterityGen, which uses RL to pretrain large-scale dexterous motion primitives, such as in-hand rotation or translation.<n>In the real world, we use human teleoperation as a prompt to the controller to produce highly dexterous behavior.
  arXiv  Detail & Related papers  (2025-02-06T18:49:35Z)
• SLIM: Sim-to-Real Legged Instructive Manipulation via Long-Horizon Visuomotor Learning [20.33419404756149]
  We present a low-cost legged mobile manipulation system that solves real-world tasks, trained by reinforcement learning purely in simulation.<n>A single policy autonomously solves long-horizon tasks involving search, move to, grasp, transport, and drop into, achieving nearly 80% real-world success.<n>This performance is comparable to that of expert human teleoperation on the same tasks while the robot is more efficient, operating at about 1.5x the speed of the teleoperation.
  arXiv  Detail & Related papers  (2025-01-17T01:32:18Z)
• FLaRe: Achieving Masterful and Adaptive Robot Policies with Large-Scale Reinforcement Learning Fine-Tuning [74.25049012472502]
  FLaRe is a large-scale Reinforcement Learning framework that integrates robust pre-trained representations, large-scale training, and gradient stabilization techniques. Our method aligns pre-trained policies towards task completion, achieving state-of-the-art (SoTA) performance on previously demonstrated and on entirely novel tasks and embodiments.
  arXiv  Detail & Related papers  (2024-09-25T03:15:17Z)
• Continual Diffuser (CoD): Mastering Continual Offline Reinforcement Learning with Experience Rehearsal [54.93261535899478]
  In real-world applications, such as robotic control of reinforcement learning, the tasks are changing, and new tasks arise in a sequential order.<n>This situation poses the new challenge of plasticity-stability trade-off for training an agent who can adapt to task changes and retain acquired knowledge.<n>We propose a rehearsal-based continual diffusion model, called Continual diffuser (CoD), to endow the diffuser with the capabilities of quick adaptation (plasticity) and lasting retention (stability)
  arXiv  Detail & Related papers  (2024-09-04T08:21:47Z)
• Collision Avoidance and Navigation for a Quadrotor Swarm Using End-to-end Deep Reinforcement Learning [8.864432196281268]
  We propose an end-to-end DRL approach to control quadrotor swarms in environments with obstacles. We provide our agents a curriculum and a replay buffer of the clipped collision episodes to improve performance in obstacle-rich environments. Our work is the first work that demonstrates the possibility of learning neighbor-avoiding and obstacle-avoiding control policies trained with end-to-end DRL.
  arXiv  Detail & Related papers  (2023-09-23T06:56:28Z)
• Generative Adversarial Imitation Learning for End-to-End Autonomous Driving on Urban Environments [0.8122270502556374]
  Generative Adversarial Imitation Learning (GAIL) can train policies without explicitly requiring to define a reward function. We show that both of them are capable of imitating the expert trajectory from start to end after training ends.
  arXiv  Detail & Related papers  (2021-10-16T15:04:13Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.