Theory of Space: Can Foundation Models Construct Spatial Beliefs through Active Exploration?
Spatial embodied intelligence requires agents to act to acquire information under partial observability. While multimodal foundation models excel at passive perception, their capacity for active, self-directed exploration remains understudied. We propose Theory of Space, defined as an agent's ability to actively acquire information through self-directed, active exploration and to construct, revise, and exploit a spatial belief from sequential, partial observations. We evaluate this through a benchmark where the goal is curiosity-driven exploration to build an accurate cognitive map. A key innovation is spatial belief probing, which prompts models to reveal their internal spatial represent...
| Authors | Pingyue Zhang, Zihan Huang, Yue Wang et al. |
|---|---|
| Primary category | cs.AI |
| Published | 4 February 2026 |
| Last updated at source | 4 February 2026 |
Connected in the knowledge graph
- MLL-Lab/Theory-of-SpaceDataset · cited by
Models and datasets referencing this paper
- MLL-Lab/Theory-of-SpaceDataset
Silytics shows public metadata from arXiv and links to the original. Information may be incomplete or out of date. Last refreshed 18 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.