Saved in:
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2602.22208 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866918357081522176 |
|---|---|
| author | Savva, Georgy Michel, Oscar Lu, Daohan Waiwitlikhit, Suppakit Meehan, Timothy Mishra, Dhairya Poddar, Srivats Lu, Jack Xie, Saining |
| author_facet | Savva, Georgy Michel, Oscar Lu, Daohan Waiwitlikhit, Suppakit Meehan, Timothy Mishra, Dhairya Poddar, Srivats Lu, Jack Xie, Saining |
| contents | Existing action-conditioned video generation models (video world models) are limited to single-agent perspectives, failing to capture the multi-agent interactions of real-world environments. We introduce Solaris, a multiplayer video world model that simulates consistent multi-view observations. To enable this, we develop a multiplayer data system designed for robust, continuous, and automated data collection on video games such as Minecraft. Unlike prior platforms built for single-player settings, our system supports coordinated multi-agent interaction and synchronized videos + actions capture. Using this system, we collect 12.64 million multiplayer frames and propose an evaluation framework for multiplayer movement, memory, grounding, building, and view consistency. We train Solaris using a staged pipeline that progressively transitions from single-player to multiplayer modeling, combining bidirectional, causal, and Self Forcing training. In the final stage, we introduce Checkpointed Self Forcing, a memory-efficient Self Forcing variant that enables a longer-horizon teacher. Results show our architecture and training design outperform existing baselines. Through open-sourcing our system and models, we hope to lay the groundwork for a new generation of multi-agent world models. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2602_22208 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Solaris: Building a Multiplayer Video World Model in Minecraft Savva, Georgy Michel, Oscar Lu, Daohan Waiwitlikhit, Suppakit Meehan, Timothy Mishra, Dhairya Poddar, Srivats Lu, Jack Xie, Saining Computer Vision and Pattern Recognition Existing action-conditioned video generation models (video world models) are limited to single-agent perspectives, failing to capture the multi-agent interactions of real-world environments. We introduce Solaris, a multiplayer video world model that simulates consistent multi-view observations. To enable this, we develop a multiplayer data system designed for robust, continuous, and automated data collection on video games such as Minecraft. Unlike prior platforms built for single-player settings, our system supports coordinated multi-agent interaction and synchronized videos + actions capture. Using this system, we collect 12.64 million multiplayer frames and propose an evaluation framework for multiplayer movement, memory, grounding, building, and view consistency. We train Solaris using a staged pipeline that progressively transitions from single-player to multiplayer modeling, combining bidirectional, causal, and Self Forcing training. In the final stage, we introduce Checkpointed Self Forcing, a memory-efficient Self Forcing variant that enables a longer-horizon teacher. Results show our architecture and training design outperform existing baselines. Through open-sourcing our system and models, we hope to lay the groundwork for a new generation of multi-agent world models. |
| title | Solaris: Building a Multiplayer Video World Model in Minecraft |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2602.22208 |