YoCausal: How Far is Video Generation from World Model? A Causality Perspective
:::info Stub — Full Engineering Breakdown Coming This paper was featured on Hugging Face Daily Papers on 2026-05-28 with 43 upvotes. A full breakdown with production viability rating, implementation notes, and honest limitations is being written. Subscribe to AI Letters → :::
| Authors | You-Zhe Xie et al. |
| Year | 2026 |
| HF Upvotes | 43 |
| arXiv | 2605.30346 |
| Download | |
| HF Page | View on Hugging Face |
Abstract
As video diffusion models (VDMs) advance toward world models, a key question arises: do they truly understand causality, or merely overfit to statistical temporal patterns? Existing benchmarks mostly rely on synthetic data, limiting real-world generalization due to the sim-to-real gap. We present YoCausal, a two-level benchmark inspired by the Violation of Expectation (VoE) paradigm from cognitive science. By temporally reversing real-world videos at zero cost as natural counterfactual samples, YoCausal establishes an arbitrarily extensible evaluation protocol. Level 1 introduces the Reverse Surprise Index (RSI), quantifying arrow-of-time perception via denoising loss. Level 2 introduces the Causality Cognition Index (CCI), which leverages a VLM to stratify datasets into causal and non-causal subsets, disentangling genuine causal reasoning from temporal bias. Evaluation of 13 state-of-the-art VDMs reveals that perceiving the arrow of time does not imply understanding causality, and a significant gap persists relative to human-level causal cognition.
Engineering Breakdown
The Problem
Existing benchmarks mostly rely on synthetic data, limiting real-world generalization due to the sim-to-real gap.
The Approach
We present YoCausal, a two-level benchmark inspired by the Violation of Expectation (VoE) paradigm from cognitive science.
Key Results
Evaluation of 13 state-of-the-art VDMs reveals that perceiving the arrow of time does not imply understanding causality, and a significant gap persists relative to human-level causal cognition.
Research Areas
This paper contributes to the following areas of AI/ML engineering:
- Machine learning
- Deep learning
- Neural networks
- Model optimization
- AI systems
- Generation
:::tip Subscribe Get weekly breakdowns of papers like this in AI Letters - the newsletter for engineers building production AI systems. :::