Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2502.01366 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866908398328479744 |
|---|---|
| author | Yin, Shaofeng Wu, Jialong Huang, Siqiao Su, Xingjian He, Xu Hao, Jianye Long, Mingsheng |
| author_facet | Yin, Shaofeng Wu, Jialong Huang, Siqiao Su, Xingjian He, Xu Hao, Jianye Long, Mingsheng |
| contents | Heterogeneity in sensors and actuators across environments poses a significant challenge to building large-scale pre-trained world models on top of this low-dimensional sensor information. In this work, we explore pre-training world models for heterogeneous environments by addressing key transfer barriers in both data diversity and model flexibility. We introduce UniTraj, a unified dataset comprising over one million trajectories from 80 environments, designed to scale data while preserving critical diversity. Additionally, we propose TrajWorld, a novel architecture capable of flexibly handling varying sensor and actuator information and capturing environment dynamics in-context. Pre-training TrajWorld on UniTraj yields substantial gains in transition prediction, achieves a new state-of-the-art for off-policy evaluation, and also delivers superior online performance of model predictive control. To the best of our knowledge, this work, for the first time, demonstrates the transfer benefits of world models across heterogeneous and complex control environments. Code and data are available at https://github.com/thuml/TrajWorld. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2502_01366 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Trajectory World Models for Heterogeneous Environments Yin, Shaofeng Wu, Jialong Huang, Siqiao Su, Xingjian He, Xu Hao, Jianye Long, Mingsheng Machine Learning Heterogeneity in sensors and actuators across environments poses a significant challenge to building large-scale pre-trained world models on top of this low-dimensional sensor information. In this work, we explore pre-training world models for heterogeneous environments by addressing key transfer barriers in both data diversity and model flexibility. We introduce UniTraj, a unified dataset comprising over one million trajectories from 80 environments, designed to scale data while preserving critical diversity. Additionally, we propose TrajWorld, a novel architecture capable of flexibly handling varying sensor and actuator information and capturing environment dynamics in-context. Pre-training TrajWorld on UniTraj yields substantial gains in transition prediction, achieves a new state-of-the-art for off-policy evaluation, and also delivers superior online performance of model predictive control. To the best of our knowledge, this work, for the first time, demonstrates the transfer benefits of world models across heterogeneous and complex control environments. Code and data are available at https://github.com/thuml/TrajWorld. |
| title | Trajectory World Models for Heterogeneous Environments |
| topic | Machine Learning |
| url | https://arxiv.org/abs/2502.01366 |