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

arXiv 26.3 2026 39.1 method

TLDR

SimDist pretrains world models in simulation and adapts them to real-world robot tasks via supervised system identification, enabling rapid improvement.

Reasoning

The paper presents a novel framework that leverages physics simulators for scalable pretraining of world models, addressing data scarcity in real-world robot learning. Its strengths include clear methodology and strong empirical results on contact-rich tasks, but it may be limited by reliance on simulator fidelity and task-specific assumptions.

Read-first score

Read-first score 39.1, weighted from topical fit, citation, graph, method, reproducibility, and recency signals. Original total remains 51.

Recency 5%

100

Uses a gentle age decay so recent papers surface without erasing older foundations. 2026

Topical relevance 25%

72.9

Uses existing LLM keyword relevance scores normalized to 0-100. world model,world simulator,generative world model,interactive world model,video world model,world dynamics prediction,model-based reinforcement learning world model

Reproducibility 15%

Screens links and visible text for paper, code, dataset, artifact, and repository signals. pdf=True; code=False; dataset=False; markers=code,github

Methodology quality 15%

Screens visible abstract and analysis fields for experiment, dataset, baseline, metric, and limitation evidence. markers=none

Citation impact 15%

14.1

Uses OpenAlex-shaped citation metadata as a bibliometric attention signal, separate from paper quality. citation_normalized_percentile=0.14095081

Graph prestige 15%

5.7

PageRank-style prestige over the local citation graph built from available OpenAlex references. local_pagerank=0.0016

Citation velocity 10%

Citation velocity estimates citations per publication-year to reduce old-paper bias. velocity=0.00

Field roles

Frontier

Rank sensitivity

Stability: volatile; rank range: 386.

Keyword Scores

world model

world dynamics prediction

model-based reinforcement learning world model

world simulator

interactive world model

generative world model

video world model

Deep Analysis

Innovations

Simulation Distillation (SimDist) framework that uses physics simulators as a scalable source of action-conditioned robot experience for pretraining world models
Transfer of encoder, reward model, and value function from simulation to real world, updating only the latent dynamics model during real-world adaptation
Reduction of real-world adaptation to supervised system identification while preserving dense, long-horizon planning signals

Methodology

SimDist pretrains a world model in simulation by distilling structural priors from a physics simulator, enabling planning from raw real-world observations. During real-world adaptation, the encoder, reward model, and value function are transferred from simulation, and only the latent dynamics model is updated using real-world prediction losses.

Key Results

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.

Resources

Paper Project