Contrastive World Models

arXiv:2609.22175v1 Announce Type: new Abstract: World models trained via pixel reconstruction can struggle in visually complex environments, where irrelevant information dominates the objective and distract the model from information relevant to planning and control. We present Contrastive World Models, an approach for learning latent dynamics models without pixel reconstruction. Building on Dreamer, we replace observation reconstruction in the standard world model objective with a Deep InfoMax-like lower bound that maximizes the mutual information between state-action sequences and local patch features of future observations, encouraging state representations to retain information that is predictive of the future without requiring the model to reconstruct visually irrelevant details. We evaluate our approach in small-scale experiments across three settings of increasing visual complexity. Our method matches Dreamer and a momentum prediction baseline in the default setting, and substantially outperforms both once distractors or natural video backgrounds are introduced, while also training more efficiently by removing the pixel decoder entirely. Our approach is general and makes minimal assumptions beyond access to state-action sequences and future observations. These results suggest that contrastive, infomax-based objectives are a principled and promising direction for building world models that are robust to visual nuisance factors, a property particularly relevant for transferring model-based RL agents to the real world.

Abstract:World models trained via pixel reconstruction can struggle in visually complex environments, where irrelevant information dominates the objective and distract the model from information relevant to planning and control. We present Contrastive World Models, an approach for learning latent dynamics models without pixel reconstruction. Building on Dreamer, we replace observation reconstruction in the standard world model objective with a Deep InfoMax-like lower bound that maximizes the mutual information between state-action sequences and local patch features of future observations, encouraging state representations to retain information that is predictive of the future without requiring the model to reconstruct visually irrelevant details. We evaluate our approach in small-scale experiments across three settings of increasing visual complexity. Our method matches Dreamer and a momentum prediction baseline in the default setting, and substantially outperforms both once distractors or natural video backgrounds are introduced, while also training more efficiently by removing the pixel decoder entirely. Our approach is general and makes minimal assumptions beyond access to state-action sequences and future observations. These results suggest that contrastive, infomax-based objectives are a principled and promising direction for building world models that are robust to visual nuisance factors, a property particularly relevant for transferring model-based RL agents to the real world.

Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV)

arXiv:2609.22175 [cs.LG]

(or arXiv:2609.22175v1 [cs.LG] for this version)

https://doi.org/10.48550/arXiv.2609.22175 Focus to learn more arXiv-issued DOI via DataCite (pending registration)

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.

aiqbench analysis

Contrastive World Models

Related coverage