Hao Shi

A single adversarial token can bypass most safety filters.

World action models do not need to render clean future video to generalize: conditioning on a single forward pass of pure noise captures virtually all the robustness benefits of multi-step diffusion at a fraction of the compute.

This work studies a route-conditioned order of unavoidable stages that every successful executor must traverse, recoverable from offline trajectories and belonging to none of them: a route-conditioned order of unavoidable stages that every successful executor must traverse.

Angular-to-Cartesian projection distortions have long bottlenecked spherical 3D perception, but explicitly modulating voxel features with spherical range-azimuth geometry outperforms top surround-view models like SurroundOcc and TPVFormer across varied wild environments.