Inferring Hierarchical Structure in Multi-Room Maze Environments
This addresses the challenge of learning and inferring environmental structure for flexible navigation in AI systems, but it appears incremental as it builds on existing hierarchical and active inference approaches.
The paper tackled the problem of inferring hierarchical structure in multi-room maze environments from pixel-based observations, and the result was a hierarchical active inference model that enables efficient exploration and goal-directed search in room-structured mini-grid environments.
Cognitive maps play a crucial role in facilitating flexible behaviour by representing spatial and conceptual relationships within an environment. The ability to learn and infer the underlying structure of the environment is crucial for effective exploration and navigation. This paper introduces a hierarchical active inference model addressing the challenge of inferring structure in the world from pixel-based observations. We propose a three-layer hierarchical model consisting of a cognitive map, an allocentric, and an egocentric world model, combining curiosity-driven exploration with goal-oriented behaviour at the different levels of reasoning from context to place to motion. This allows for efficient exploration and goal-directed search in room-structured mini-grid environments.