Learning Implicit Causal World Models from Multi-Agent Demonstrations
This paper introduces Implicit Causal World Models to recover environmental dynamics from offline demonstrations in multi-agent systems, providing interpretable causal representations with scalable accuracy.
Introduces Implicit Causal World Models for recovering environmental dynamics in multi-agent systems
Keywords
Before reading this…
Applications
- →Coordination tasks such as Two-Door, Navigation, and Giveway
To understand this paper, make sure you know these concepts first:
- Understanding of reinforcement learning, multi-agent systems, and world modelsfind papers →
Abstract
More Like ThisIn model-based reinforcement learning, world models exist as internal simulators, but their training often conflates statistical correlations with causal mechanisms. This problem is exacerbated in multi-agent systems where physical transitions are intertwined with strategic agent intents, causing world models to fail under distribution shift. We introduce Implicit Causal World Models to recover environmental dynamics from offline demonstrations without requiring pre-defined causal graphs. By incorporating policy variance, we render world models discoverable via the sequential backdoor condition. Evaluations across coordination tasks (Two-Door, Navigation, and Giveway) demonstrate that these models provide interpretable causal representations under both full and partial observability, with model accuracy scaling directly with interventional strength.