Search papers, labs, and topics across Lattice.
The paper introduces MAS (Multiplayer world models with Authoritative Shared State), which addresses the limitations of current video world models in multiplayer settings by disentangling world dynamics from view-dependent visual latents. This approach employs a learned Logic Engine to manage a global authoritative state based on joint actions, eliminating the need for hand-written transition functions, while a Rendering Engine generates consistent views for any camera on demand. The results demonstrate that MAS significantly improves state accuracy and reduces cross-view inconsistency, achieving scalability with 1,024 concurrent players over 10,000 recurrent steps in a multiplayer Snake benchmark.
Explicitly disentangling world dynamics from view rendering allows MAS to achieve unprecedented scalability and consistency in multiplayer simulations.
Current video world models struggle in multiplayer environments because they entangle world state with view-dependent visual latents, leading to redundant compute, view inconsistencies, and poor scalability. We propose MAS (Multiplayer world models with Authoritative Shared State) to resolve this limitation. Inspired by multiplayer game architectures, MAS disentangles world dynamics and view rendering. A learned Logic Engine advances a global, authoritative typed state from joint actions without any hand-written transition function, acting as the sole recurrent memory and synchronization reference. From this shared state, a learned Rendering Engine generates independent and consistent views for any requested camera on demand. This explicit disentangling allows MAS to achieve superior state accuracy and lower cross-view inconsistency compared to state-of-the-art multi-view baselines on a matched multiplayer Snake benchmark. It advances predicted worlds with 1,024 concurrent players for 10,000 recurrent steps. Our results show that explicit, authoritative state modeling provides a practical foundation for scalable and consistent multi-agent world simulation.
