SOURCE-LINKED INTELLIGENCE

GAE: Learning a Geometry-Native Latent Space for 3D-Consistent World Generation

arXiv · AI, language, vision and robotics · article · Sep 21, 2026 · UTC

We present a compact geometry-native latent space as a shared foundation for perception and generation. Visual generators can produce photorealistic frames without preserving a consistent 3D scene. We argue that this is not only a modeling problem but also a representation problem: generators typically evolve appearance-centric latents, while perception models recover geometry in a semantically rich space that encodes cross-view structure. Rather than adding geometry as another output, we reparameterize a geometry foundation model's features into a compact latent space for generation. We reali

Read original source ↗ Open in workspace

recordType

paper

region

Global