AI RESEARCH

MemRoPE: Training-Free Infinite Video Generation via Evolving Memory Tokens

arXiv CS.CV

ArXi:2603.12513v1 Announce Type: new Autoregressive diffusion enables real-time frame streaming, yet existing sliding-window caches discard past context, causing fidelity degradation, identity drift, and motion stagnation over long horizons. Current approaches preserve a fixed set of early tokens as attention sinks, but this static anchor cannot reflect the evolving content of a growing video. We