Simulation Distillation: Pretraining World Models in Simulation for Rapid Real-World Adaptation

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Levy, Jacob, Westenbroek, Tyler, Huang, Kevin, Palafox, Fernando, Yin, Patrick, Omidshafiei, Shayegan, Kim, Dong-Ki, Gupta, Abhishek, Fridovich-Keil, David
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909037280362496
author Levy, Jacob
Westenbroek, Tyler
Huang, Kevin
Palafox, Fernando
Yin, Patrick
Omidshafiei, Shayegan
Kim, Dong-Ki
Gupta, Abhishek
Fridovich-Keil, David
author_facet Levy, Jacob
Westenbroek, Tyler
Huang, Kevin
Palafox, Fernando
Yin, Patrick
Omidshafiei, Shayegan
Kim, Dong-Ki
Gupta, Abhishek
Fridovich-Keil, David
contents Robot learning requires adaptation methods that improve reliably from limited, mixed-quality interaction data. This is especially challenging in long-horizon, contact-rich tasks, where end-to-end policy finetuning remains inefficient and brittle. World models offer a compelling alternative: by predicting the outcomes of candidate action sequences, they enable online planning through counterfactual reasoning. However, training action-conditioned robotic world models directly in the real world requires diverse data at impractical scale. We introduce Simulation Distillation (SimDist), a framework that uses physics simulators as a scalable source of action-conditioned robot experience. During pretraining, SimDist distills structural priors from the simulator into a world model that enables planning from raw real-world observations. During real-world adaptation, SimDist transfers the encoder, reward model, and value function learned in simulation, and updates only the latent dynamics model using real-world prediction losses. This reduces adaptation to supervised system identification while preserving dense, long-horizon planning signals for online improvement. Across contact-rich manipulation and quadruped locomotion tasks, SimDist rapidly improves with experience, while prior adaptation methods struggle to make progress or degrade during online finetuning. Project website and code: https://sim-dist.github.io
format Preprint
id arxiv_https___arxiv_org_abs_2603_15759
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Simulation Distillation: Pretraining World Models in Simulation for Rapid Real-World Adaptation
Levy, Jacob
Westenbroek, Tyler
Huang, Kevin
Palafox, Fernando
Yin, Patrick
Omidshafiei, Shayegan
Kim, Dong-Ki
Gupta, Abhishek
Fridovich-Keil, David
Robotics
Artificial Intelligence
Machine Learning
Robot learning requires adaptation methods that improve reliably from limited, mixed-quality interaction data. This is especially challenging in long-horizon, contact-rich tasks, where end-to-end policy finetuning remains inefficient and brittle. World models offer a compelling alternative: by predicting the outcomes of candidate action sequences, they enable online planning through counterfactual reasoning. However, training action-conditioned robotic world models directly in the real world requires diverse data at impractical scale. We introduce Simulation Distillation (SimDist), a framework that uses physics simulators as a scalable source of action-conditioned robot experience. During pretraining, SimDist distills structural priors from the simulator into a world model that enables planning from raw real-world observations. During real-world adaptation, SimDist transfers the encoder, reward model, and value function learned in simulation, and updates only the latent dynamics model using real-world prediction losses. This reduces adaptation to supervised system identification while preserving dense, long-horizon planning signals for online improvement. Across contact-rich manipulation and quadruped locomotion tasks, SimDist rapidly improves with experience, while prior adaptation methods struggle to make progress or degrade during online finetuning. Project website and code: https://sim-dist.github.io
title Simulation Distillation: Pretraining World Models in Simulation for Rapid Real-World Adaptation
topic Robotics
Artificial Intelligence
Machine Learning
url https://arxiv.org/abs/2603.15759