World in World introduces a training‑free, inference‑time interface that transforms diverse control signals—such as source‑video observations, target‑view projections, geometry renderings, and retrieved states—into camera‑ and time‑labelled visual states. These states are processed by a frozen causal video model’s self‑attention, enabling tasks like camera‑controlled rerendering, long‑horizon revisiting, and human‑motion transfer without additional training. The method employs a correspondence router and evidence‑wise attention to align token identities and regulate auxiliary channel contributions during a single denoising pass.
By Chenxi Song, Yanming Yang, Chi Zhang
arXiv:2610.33169v1 Announce Type: new
Abstract: Interactive video world models require coordinated control over viewing perspectives, character actions, and camera motion. Existing methods typically...
By Harold Haodong Chen, Rui Peng, Rongjin Guo, Lingteng Qiu, Kejie Qiu, Chao Xu, Zilong Dong
Momentum introduces a two-stage pipeline for video world models that uses synchronized multi-camera rigs to represent evolving scenes. The WorldUpdate stage generates shared reference-camera videos as a memory of scene dynamics, while WorldInteract grounds target camera video generation in this memory, ensuring moving objects remain represented outside the field of view. Experiments on CARLA street scenes show a 126.2% improvement in dynamic scene consistency over static image conditioning.
By Nedko Savov, Stefan Maria Ailuro, Danda Pani Paudel, Yilun Du, Xi Wang, Luc Van Gool
arXiv:2610.33589v1 Announce Type: new
Abstract: While interactive video world models require autoregressive generation, directly adapting pretrained bidirectional video diffusion models into autoregr...
By Zelin Gao, Hao Ouyang, Jiapeng Zhu, Qiuyu Wang, Jingye Chen, Zichen Liu, Xing Zhu, Yujun Shen, Yinghao Xu
Video world models enable interactive exploration of dynamic environments, yet struggle to respect prior observations over long horizons and across viewpoints. We present WorldCrafter, a video world m...
arXiv:2610.17434v1 Announce Type: cross
Abstract: Recent advances in video world models have enabled plausible interactive environments, yet long-horizon generation places a fundamental demand on vis...
By Junyi Wu, Zekun Li, Chenhongyi Yang, Chen Guo, Lingchen Yang, Yongan Zhang, Xinyi Zhang, Hongbing Hu, Bartek Rymkowski, Srinath Sridhar, Yan Yan, Xijie Huang
WorldCrafter is a video world model that introduces a camera‑queryable implicit 3D‑aware memory to improve long‑horizon consistency and viewpoint control. The model compresses multi‑view evidence into a limited token budget shaped by the requested viewpoint, integrating historical observations via a memory encoder and pose‑conditioned readout before denoising. Experiments on static and dynamic scenes demonstrate significant gains in consistency and camera‑control accuracy while maintaining visual quality during minute‑scale exploration.
By Wangbo Yu, Kunhao Liu, Wenbo Hu, Shenghai Yuan, Chaoran Feng, Haiyang Zhou, Yukun Huang, Yiran Wang, Wang Zhao, Yingmin Luo, Ying Shan
StarWM introduces a self‑supervised attention routing mechanism that selectively applies reconstruction only to dynamically relevant regions of visual input. By combining a cross‑attention module with a dual‑stream decoder and stop‑gradient barriers, it balances faithful environmental dynamics capture with abstraction of irrelevant content. Experiments on DeepMind Control show that StarWM outperforms both reconstruction‑based and reconstruction‑free baselines, especially under distractor conditions, and preserves state attributes over long‑horizon imagination.
By Zeqiang Zhang, Fabian Wurzberger, Maximilian Otte, Daniel Schmid, Sebastian Gottwald, Arne Peter Raulf, Daniel Alexander Braun
WorldCraft is an autoregressive video world model that enables both camera navigation and persistent object-trajectory control in interactive video environments. It introduces Normalized World Trajectory (NWT) to anchor user-specified 2D motion paths in a world coordinate system, and Trajectory-Anchored State Persistence (TASP) to maintain updated object states even when the target is temporarily invisible. The model also employs a pathway-selective LoRA to add object control while preserving pretrained camera navigation capabilities, achieving accurate composable control and maintaining object states across extended off-screen intervals.
By Bohai Gu, Taiyi Wu, Yueyang Yuan, Jian Liu, Xiaocheng Lu, Dazhao Du, Jie Zhang, Jinxiang Lai, Shuai Yang, Xiaotong Zhao, Alan Zhao, Song Guo
The paper introduces a decoupled memory control framework for spatially consistent long‑horizon video generation. By separating memory conditioning from generation, it uses a lightweight memory branch to learn spatial consistency from past observations and a per‑frame cross‑attention mechanism to condition each frame on the most relevant historical data. A camera‑aware gating mechanism further ensures memory is used only when meaningful references exist, resulting in lower training costs, higher spatial consistency, and state‑of‑the‑art visual quality.
By Yanjun Guo, Zhengqiang Zhang, Pengfei Wang, Xinyue Liang, Zhiyuan Ma, Lei Zhang
World Observer introduces a joint actor‑observer framework for persistent world modeling, decoupling the agent’s view from panoramic observers that track regions outside the actor’s field of view. By warping from a shared panoramic source and using an Observer Sink for high‑resolution references, the method maintains accurate object states and dynamics even when they leave and re‑enter the actor’s view. The approach is evaluated with new world‑space metrics and a benchmark across real and synthetic scenes, showing improved out‑of‑view dynamics while preserving visual fidelity, camera control, and 3D adherence.
By Hyunwook Choi, Dahyun Chung, Hyunsung Kim, Siyoon Jin, Jinhyeok Choi, Junyoung Seo, Seungryong Kim
SolarWM is an open foundation for building interactive video world models, offering a reconfigurable multi‑source data engine that unifies 1.43 million clips from 10 datasets into a consistent, frame‑aligned format. It provides a backbone‑native adaptation framework that preserves native representations of models ranging from 5 B to 33 B parameters, and a three‑stage training recipe combining bidirectional adaptation, teacher‑forced autoregressive initialization, and distribution‑matching distillation. The resulting causal models can interact in real‑time over rollouts from minutes to hours, trained only on 5‑second sequences, and the project releases data, pipeline, recipes, weights, and framework for reproducible research.
By Junchao Huang, Guian Fang, Shengju Qian, Xianghao Kong, Zhuoran Zhao, Wei Huang, Yihua Du, Zixin Zhang, Justin Cui, Yuchao Gu, Yukang Chen, Xinting Hu, Tianyu He, Shaoshuai Shi, Zhuotao Tian, Xin Wang, Mike Zheng Shou, Li Jiang