titleWorldDraft: If Code Builds the World, Can Video Models Just Render It?
authorsZhiyu Hou, Wen-Jie Shu, Harold Haodong Chen, Weilin Ruan, Chenfei Liao, Liuzhou Zhang, Disen Lan, Xiaogang Xu, Qi Chen, Ser-Nam Lim
date2026-11-12
venuearXiv (preprint, not peer reviewed)
idarXiv:2611.16217
upvotes16 (Hugging Face Daily Papers, 2026-11-15)
linksabs pdf hf code
cite
abstractMultimodal coding agents decouple 4D world construction from visual generation, repurposing video models from world generators toward generative renderers. This division of labor raises a fundamental question: if code builds the world, can video models just render it? A reliable generative renderer hinges on two complementary criteria: (i) world preservation, retaining prescribed 4D kinematics and camera paths; and (ii) visual realization, persistently binding text-specified target attributes over time. To evaluate this capability, we introduce WorldDraft, a benchmark of 432 strictly paired test cases across 48 executable 4D world families spanning six dynamic challenge categories. Under a 3 times 3 factorial design, WorldDraft orthogonally crosses draft granularity (abstract proxy to previs) and appearance alignment (neutral to conflicting) under identical dynamics. Extensive evaluations across 12 representative models reveal that: (I) world preservation and visual realization remain in stark operational tension; (II) richer draft scaffolding anchors structure but paradoxically impairs appearance fidelity; and (III) conflicting source cues trigger severe source leakage (up to 64%), exposing fragile cross-modal disentanglement. To mitigate these bottlenecks, we propose WorldDraft-Primer, a plug-and-play visual prompting strategy that consistently improves both objectives across frozen video renderers. Together, WorldDraft and WorldDraft-Primer lay a foundation for bridging programmatic 4D simulation with video generation, advancing physically grounded generative rendering.

bibtex

@misc{hou2026worlddraft,
  title={WorldDraft: If Code Builds the World, Can Video Models Just Render It?},
  author={Zhiyu Hou and Wen-Jie Shu and Harold Haodong Chen and Weilin Ruan and Chenfei Liao and Liuzhou Zhang and Disen Lan and Xiaogang Xu and Qi Chen and Ser-Nam Lim},
  year={2026},
  eprint={2611.16217},
  archivePrefix={arXiv},
  url={https://arxiv.org/abs/2611.16217}
}