# Disen-Forcing: Disentangling Semantic Anchoring from Motion for Autoregressive Video Diffusion

> Source: <https://aiflash.com/news/152278/>
> Published: 2026-11-16 05:30:04+00:00

Autoregressive video diffusion models (AR-VDMs) enable real-time, long-horizon video generation but degrade rapidly once the inference horizon outruns the training horizon, due to compounding exposure bias. To stabilize long rollouts, recent methods heuristically adopt frame sink, keeping initial fr
