Chi Zhang

Uncertainty DMD is proposed, a simple uncertainty-injection framework that restores stochasticity at two key stages of AR generation: a timestep perturbation for the first chunk to increase first-chunk diversity, and a stochastic cache-writing mechanism for later chunks to preserve uncertainty in autoregressive conditioning.

World in World is presented, a training-free inference-time interface that converts heterogeneous control evidence into camera- and time-labelled clean visual states, which are read through the native self attention of a frozen causal video model.

Unified multimodal models can bootstrap both visual generation and perceptual understanding purely from text prompts by treating their own generative flow trajectories as self-supervised grounding signals.

Achieving superior 3D scene reconstruction from a single measurement, this method combines Gaussian Splatting with advanced vision model priors to tackle the complexities of Snapshot Compressive Imaging.