NitroGen: An Open Foundation Model for Generalist Gaming Agents
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , , , , , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2026
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866914234807353344 |
|---|---|
| author | Magne, Loïc Awadalla, Anas Wang, Guanzhi Xu, Yinzhen Belofsky, Joshua Hu, Fengyuan Kim, Joohwan Schmidt, Ludwig Gkioxari, Georgia Kautz, Jan Yue, Yisong Choi, Yejin Zhu, Yuke Fan, Linxi "Jim" |
| author_facet | Magne, Loïc Awadalla, Anas Wang, Guanzhi Xu, Yinzhen Belofsky, Joshua Hu, Fengyuan Kim, Joohwan Schmidt, Ludwig Gkioxari, Georgia Kautz, Jan Yue, Yisong Choi, Yejin Zhu, Yuke Fan, Linxi "Jim" |
| contents | We introduce NitroGen, a vision-action foundation model for generalist gaming agents that is trained on 40,000 hours of gameplay videos across more than 1,000 games. We incorporate three key ingredients: 1) an internet-scale video-action dataset constructed by automatically extracting player actions from publicly available gameplay videos, 2) a multi-game benchmark environment that can measure cross-game generalization, and 3) a unified vision-action model trained with large-scale behavior cloning. NitroGen exhibits strong competence across diverse domains, including combat encounters in 3D action games, high-precision control in 2D platformers, and exploration in procedurally generated worlds. It transfers effectively to unseen games, achieving up to 52% relative improvement in task success rates over models trained from scratch. We release the dataset, evaluation suite, and model weights to advance research on generalist embodied agents. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2601_02427 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | NitroGen: An Open Foundation Model for Generalist Gaming Agents Magne, Loïc Awadalla, Anas Wang, Guanzhi Xu, Yinzhen Belofsky, Joshua Hu, Fengyuan Kim, Joohwan Schmidt, Ludwig Gkioxari, Georgia Kautz, Jan Yue, Yisong Choi, Yejin Zhu, Yuke Fan, Linxi "Jim" Computer Vision and Pattern Recognition Artificial Intelligence Machine Learning We introduce NitroGen, a vision-action foundation model for generalist gaming agents that is trained on 40,000 hours of gameplay videos across more than 1,000 games. We incorporate three key ingredients: 1) an internet-scale video-action dataset constructed by automatically extracting player actions from publicly available gameplay videos, 2) a multi-game benchmark environment that can measure cross-game generalization, and 3) a unified vision-action model trained with large-scale behavior cloning. NitroGen exhibits strong competence across diverse domains, including combat encounters in 3D action games, high-precision control in 2D platformers, and exploration in procedurally generated worlds. It transfers effectively to unseen games, achieving up to 52% relative improvement in task success rates over models trained from scratch. We release the dataset, evaluation suite, and model weights to advance research on generalist embodied agents. |
| title | NitroGen: An Open Foundation Model for Generalist Gaming Agents |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence Machine Learning |
| url | https://arxiv.org/abs/2601.02427 |