Search papers, labs, and topics across Lattice.
5
0
8
9
Imagined visual evidence can dramatically enhance spatial reasoning in VLMs, leading to significant performance gains on complex reasoning tasks.
Ditch the VAE bottleneck: Representation Forcing lets you train unified multimodal models to generate high-quality images directly from pixels, rivaling VAE-based approaches without the architectural constraint.
Interactive 3D asset generation can now be driven by functional logic and hierarchical physics, thanks to a new framework that synthesizes simulation-ready assets.
Accurately simulating multi-agent interactions with consistent multi-view video is now possible thanks to MultiWorld, a framework that scales to many agents and viewpoints.
EgoSim delivers spatially consistent and dynamically updating egocentric simulations, outperforming existing methods and enabling cross-embodiment transfer to robotic manipulation.
