---
title: "Microsoft Research Finds Hidden Physics Inside Video Diffusion Models"
subtitle: "A new paper shows video diffusion models secretly encode physical laws in their transformer layers — despite never being trained to do so"
url: "https://alphasignal.ai/news/microsoft-research-finds-hidden-physics-inside-video-diffusion-models"
author: "AlphaSignal Newsroom"
author_url: "https://alphasignal.ai/editorial-team"
company: "AlphaSignal"
company_url: "https://alphasignal.ai/company/alphasignal"
published: "2026-06-03T00:00:00.000Z"
updated: "2026-06-11T03:54:38.472Z"
type: "Paper"
topics:
  - "Video"
  - "Reasoning"
subtopics:
  - "World Models"
  - "Video Generation"
  - "Test Time Compute"
upvotes: 1459
source_url: "https://arxiv.org/abs/2606.05328"
comments: 0
truncated: true
---

# Microsoft Research Finds Hidden Physics Inside Video Diffusion Models

> A new paper shows video diffusion models secretly encode physical laws in their transformer layers — despite never being trained to do so

Video diffusion models like Sora, Veo, and Cosmos are increasingly pitched as world simulators. But a nagging question has followed that ambition: do these models actually _understand_ physics, or are they just very good at copying motion patterns from training data? A new paper from researchers at the University of Bristol, McGill, Mila, and Microsoft Research has a surprising answer. The physics is in there , and it is more accessible than anyone expected.

The paper, ["The Invisible Hand of Physics"](https://arxiv.org/abs/2606.05328), probes the internal states of three large-scale video diffusion models and finds that physical plausibility is **linearly decodable** from their transformer layers. That means a simple linear classifier , no fine-tuning, no complex head , can read off whether a video obeys the laws of physics directly from the model's hidden activations. The catch: this signal is completely absent from the model's input. It emerges inside the transformer during the denoising computation itself.

## Why this question is hard to answer

Modern video diffusion models operate in the latent space of a variational autoencoder (VAE) , a compression network trained purely to reconstruct video pixels. That VAE has no reason to organize its latent space around physics. And the diffusion model on top of it is trained only to denoise: to gradually transform random noise into a clean video. There is no explicit physics objective anywhere in the pipeline.

The other problem is access. Diffusion models generate videos by starting from noise and integrating forward. They never "see" a real video as an input , they only produce outputs. So you cannot simply feed a real video through the model and inspect what it thinks about it. The internal states the researchers wanted to probe simply do not exist in normal inference.

The prior state of the art for physics understanding in video models came from a different family entirely: self-supervised predictive encoders like **V-JEPA** (Meta's video joint embedding predictive architecture). V-JEPA, trained to predict future video states in representation space, achieves 98% zero-shot accuracy on IntPhys and 62% on InfLevel , benchmarks designed to test whether a model can distinguish physically plausible from physically impossible events. The hypothesis was that a predictive training objective was the key ingredient for physics understanding to emerge.

## The trick: running the model backwards

The key methodological innovation is **reverse sampling**. Instead of generating from noise, the researchers take a real video, encode it with the VAE to get a clean latent, and then integrate the model's learned velocity field

_Preview only. Full text requires an AlphaSignal Pro account._

## Sources

- [The Invisible Hand of Physics: When Video Diffusion Models Know More Than They Show](https://arxiv.org/abs/2606.05328)
- [Intuitive physics understanding emerges from self-supervised pretraining on natural videos (Garrido et al., 2025)](https://arxiv.org/html/2502.11831v1)
- [IntPhys: A Benchmark for Visual Intuitive Physics Reasoning](https://intphys.cognitive-ml.fr/)
- [GitHub: videodiffusionphysics](https://github.com/ParsaEsmati/videodiffusionphysics)