![[2602.23152] The Trinity of Consistency as a Defining Principle for General World Models](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2602.23152] The Trinity of Consistency as a Defining Principle for General World Models
Summary
This paper proposes the 'Trinity of Consistency' as a foundational principle for developing General World Models in AI, emphasizing modal, spatial, and temporal consistency.
Why It Matters
The establishment of a theoretical framework for General World Models is crucial for advancing Artificial General Intelligence. By addressing the limitations of current models and proposing a structured approach, this work could guide future research and development in AI, particularly in multimodal learning and reasoning.
Key Takeaways
- Introduces the Trinity of Consistency: modal, spatial, and temporal.
- Highlights the evolution of multimodal learning towards unified architectures.
- Presents CoW-Bench, a benchmark for evaluating video generation models.
- Clarifies limitations of current systems and architectural needs for future progress.
- Aims to provide a principled pathway for developing General World Models.
Computer Science > Artificial Intelligence arXiv:2602.23152 (cs) [Submitted on 26 Feb 2026] Title:The Trinity of Consistency as a Defining Principle for General World Models Authors:Jingxuan Wei, Siyuan Li, Yuhang Xu, Zheng Sun, Junjie Jiang, Hexuan Jin, Caijun Jia, Honghao He, Xinglong Xu, Xi bai, Chang Yu, Yumou Liu, Junnan Zhu, Xuanhe Zhou, Jintao Chen, Xiaobin Hu, Shancheng Pang, Bihui Yu, Ran He, Zhen Lei, Stan Z. Li, Conghui He, Shuicheng Yan, Cheng Tan View a PDF of the paper titled The Trinity of Consistency as a Defining Principle for General World Models, by Jingxuan Wei and 23 other authors View PDF Abstract:The construction of World Models capable of learning, simulating, and reasoning about objective physical laws constitutes a foundational challenge in the pursuit of Artificial General Intelligence. Recent advancements represented by video generation models like Sora have demonstrated the potential of data-driven scaling laws to approximate physical dynamics, while the emerging Unified Multimodal Model (UMM) offers a promising architectural paradigm for integrating perception, language, and reasoning. Despite these advances, the field still lacks a principled theoretical framework that defines the essential properties requisite for a General World Model. In this paper, we propose that a World Model must be grounded in the Trinity of Consistency: Modal Consistency as the semantic interface, Spatial Consistency as the geometric basis, and Temporal Consistency a...
