| title | PrimeSim: A Generalist World Model Simulator from Large-Scale Cross-Embodiment Demonstrations |
|---|---|
| authors | Travis Davies, Wentao Ma, Lihao Guo, Tengfei Ma, Xiaowen Chang, Jiafeng Xu, Shilong Wang, Yiqi Deng, Shaowen He, Xiaoqi Li, Qian Zhu, Ran Cheng, Hao Dong |
| date | 2026-11-03 |
| venue | arXiv (preprint, not peer reviewed) |
| id | arXiv:2611.04870 |
| upvotes | 9 (Hugging Face Daily Papers, 2026-11-05) |
| links | abs pdf hf |
| cite | |
| primer | World models |
| abstract | A robot world model that serves as a simulator must do more than generate plausible videos: it must faithfully follow commanded actions, maintain consistency across camera views, and sustain coherent long-horizon rollouts. We present PrimeSim, an action-conditioned, multiview world model simulator that addresses these challenges through progressive midtraining and posttraining. Starting from a pretrained video generator, we midtrain on 30,110 hours of robot demonstrations spanning 22 platforms. To improve action following, we introduce a unified cross-embodiment action representation and frame-aligned dual-path conditioning to strengthen action-frame correspondence. To maintain multiview consistency, we develop view-aware interleaved attention with explicit camera identities, enabling cross-view information exchange while preserving view-specific representations. We further introduce a fast-video, slow-memory posttraining framework that transforms the bidirectional video generator into an efficient autoregressive simulator. By combining fast causal generation, self-rollout training, and hierarchical temporal memory, PrimeSim mitigates error accumulation from self-generated histories and supports coherent long-horizon simulation. PrimeSim runs at 10.1 FPS, remains stable over long-horizon rollouts on rigid and deformable manipulation tasks, and transfers to held-out platforms within 100 to 200 gradient steps where the pretrained backbone does not. Against five open-source action-conditioned world models it leads four of the five WorldSimProbe suites and posts the best overall EWMBench score, while fine-grained local calibration still trails the field; its rollouts are faithful enough to support GRPO-based policy optimization. We release the model weights, inference code, and training code. |
bibtex
@misc{davies2026primesim,
title={PrimeSim: A Generalist World Model Simulator from Large-Scale Cross-Embodiment Demonstrations},
author={Travis Davies and Wentao Ma and Lihao Guo and Tengfei Ma and Xiaowen Chang and Jiafeng Xu and Shilong Wang and Yiqi Deng and Shaowen He and Xiaoqi Li and Qian Zhu and Ran Cheng and Hao Dong},
year={2026},
eprint={2611.04870},
archivePrefix={arXiv},
url={https://arxiv.org/abs/2611.04870}
}