SEAL-Bench

SEAL-Bench is a benchmark for evaluating spatial reasoning in AI models, developed by the research community. It measures accuracy on spatial understanding tasks, with SpatialTree achieving 79.8% accuracy compared to GPT-4V's 67.4%.

Two-hop subgraph: this entity, every entity it directly relates to, and every entity those neighbors relate to. Drag a node, scroll to zoom, click to inspect — or click any neighbor and re-center the atlas there.