Simple, Good, Fast: Self-Supervised World Models Free of Baggage
Machine Learning
2025-06-04 v1 Artificial Intelligence
Machine Learning
Abstract
What are the essential components of world models? How far do we get with world models that are not employing RNNs, transformers, discrete representations, and image reconstructions? This paper introduces SGF, a Simple, Good, and Fast world model that uses self-supervised representation learning, captures short-time dependencies through frame and action stacking, and enhances robustness against model errors through data augmentation. We extensively discuss SGF's connections to established world models, evaluate the building blocks in ablation studies, and demonstrate good performance through quantitative comparisons on the Atari 100k benchmark.
Keywords
Cite
@article{arxiv.2506.02612,
title = {Simple, Good, Fast: Self-Supervised World Models Free of Baggage},
author = {Jan Robine and Marc Höftmann and Stefan Harmeling},
journal= {arXiv preprint arXiv:2506.02612},
year = {2025}
}
Comments
Published as a conference paper at ICLR 2025. Code is available at https://github.com/jrobine/sgf