Predictive world models that simulate future observations under explicit camera control are fundamental to interactive AI. Despite rapid advances, current systems lack spatial persistence: they fail to maintain stable scene structures over long trajectories, frequently hallucinating details when cameras revisit previously observed locations. We identify that this geometric drift stems from reliance on screen-space positional embeddings, which conflict with the projective geometry required for 3D consistency. We introduce \textbf{ViewRope}, a geometry-aware encoding that injects camera-ray directions directly into video transformer self-attention layers. By parameterizing attention with relative ray geometry rather than pixel locality, ViewRope provides a model-native inductive bias for retrieving 3D-consistent content across temporal gaps. We further propose \textbf{Geometry-Aware Frame-Sparse Attention}, which exploits these geometric cues to selectively attend to relevant historical frames, improving efficiency without sacrificing memory consistency. We also present \textbf{ViewBench}, a diagnostic suite measuring loop-closure fidelity and geometric drift. Our results demonstrate that ViewRope substantially improves long-term consistency while reducing computational costs.
@article{arxiv.2602.07854,
title = {Geometry-Aware Rotary Position Embedding for Consistent Video World Model},
author = {Chendong Xiang and Jiajun Liu and Jintao Zhang and Xiao Yang and Zhengwei Fang and Shizun Wang and Zijun Wang and Yingtian Zou and Hang Su and Jun Zhu},
journal= {arXiv preprint arXiv:2602.07854},
year = {2026}
}