Learning Massively Multitask World Models for Continuous Control
Hansen, Nicklas, Su, Hao, Wang, Xiaolong
–arXiv.org Artificial Intelligence
General-purpose control demands agents that act across many tasks and embodiments, yet research on reinforcement learning (RL) for continuous control remains dominated by single-task or offline regimes, reinforcing a view that online RL does not scale. Inspired by the foundation model recipe (large-scale pretraining followed by light RL) we ask whether a single agent can be trained on hundreds of tasks with online interaction. To accelerate research in this direction, we introduce a new benchmark with 200 diverse tasks spanning many domains and embodiments, each with language instructions, demonstrations, and optionally image observations. We then present \emph{Newt}, a language-conditioned multitask world model that is first pretrained on demonstrations to acquire task-aware representations and action priors, and then jointly optimized with online interaction across all tasks. Experiments show that Newt yields better multitask performance and data-efficiency than a set of strong baselines, exhibits strong open-loop control, and enables rapid adaptation to unseen tasks. We release our environments, demonstrations, code for training and evaluation, as well as 200+ checkpoints.
arXiv.org Artificial Intelligence
Dec-3-2025
- Country:
- Asia > Japan
- Honshū > Chūbu > Ishikawa Prefecture > Kanazawa (0.04)
- Europe
- Netherlands > South Holland
- Delft (0.04)
- Portugal > Braga
- Braga (0.04)
- Netherlands > South Holland
- North America > United States
- California > San Diego County
- San Diego (0.04)
- Indiana (0.04)
- California > San Diego County
- Asia > Japan
- Genre:
- Research Report > New Finding (0.67)
- Industry:
- Leisure & Entertainment > Games > Computer Games (0.93)
- Technology:
- Information Technology > Artificial Intelligence
- Cognitive Science (1.00)
- Machine Learning
- Neural Networks > Deep Learning (0.68)
- Reinforcement Learning (1.00)
- Natural Language (1.00)
- Representation & Reasoning (1.00)
- Robots (1.00)
- Information Technology > Artificial Intelligence