Physics IQ Benchmark:
Do generative video models learn physical principles from watching videos?

TL;DR

We develop the Physics-IQ benchmark and score, which reveals that current generative video models lack physical understanding despite sometimes achieving visual realism. Use our benchmark and dataset to assess your video model's physics understanding!

Abstract 📣

AI video generation is undergoing a revolution, with quality and realism advancing rapidly. These advances have led to a passionate scientific debate: Do video models learn “world models” that discover laws of physics — or, alternatively, are they merely sophisticated pixel predictors that achieve visual realism without understanding the physical principles of reality?

We address this question by developing Physics-IQ, a comprehensive benchmark dataset that can only be solved by acquiring a deep understanding of various physical principles, like fluid dynamics, optics, solid mechanics, magnetism and thermodynamics. We find that across a range of current models (Sora, Runway, Pika, Lumiere, Stable Video Diffusion, and VideoPoet), physical understanding is severely limited. At the same time, some test cases can already be successfully solved. For instance, fluid dynamics tend to work better than solid mechanics. This indicates that acquiring certain physical principles from observation alone may be possible, but significant challenges remain. While we expect rapid advances ahead, our work demonstrates that visual realism does not imply physical understanding.

Examples 🎞️