Probing the Impact of Scale on Data-Efficient, Generalist Transformer World Models for Atari
For researchers building data-efficient generalist agents, this work clarifies the role of scale in world models, showing that joint training can overcome environment-specific scaling limitations.
This paper investigates how model scale impacts world model performance in Atari 100k, finding that environments fall into distinct scaling regimes, but joint training across 26 environments stabilizes scaling and yields monotonic gains. The best world model achieves a median expert-random-normalized score of 0.770 in downstream control.
Developing generalist systems that retain human-like data efficiency is a central challenge. While world models (WMs) offer a promising path, existing research often conflates architectural mechanisms with the independent impact of model \emph{scale}. In this work, we use a minimalist transformer world model to analyze scaling behaviors on the Atari 100k benchmark, using fixed offline datasets derived from a presupposed expert policy. Our results reveal that environments fundamentally fall into distinct scaling regimes, even when constrained by identical offline data budgets and model capacities. For individual tasks, some environments naturally allow models to pass the interpolation threshold, yielding monotonic improvements in the overparameterized regime, while others remain trapped in the classical regime, where larger world models degrade fidelity. In the unified setting, i.e., a single transformer trained on a suite of 26 Atari environments, we uncover that joint training stabilizes scaling dynamics, ensuring monotonic gains across all environments, regardless of their distinct inherent scaling regimes. Finally, we demonstrate that improved fidelity translates directly to downstream control, with policies learned entirely within the simulated dynamics achieving a median expert-random-normalized score of 0.770. Our findings suggest that future progress lies as much in precise scaling strategies as in architectural innovation.