# PhysWeep: Does a Video Generator Realize the Physics You Ask For?

arXiv: https://arxiv.org/abs/2609.06207 · HTML: https://mlsanity.com/paper/2609.06207/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Builds on 31 of its 38 references that are in the map.
- First posted on arXiv 5 Sep 2026.

## Builds on (most cited first, 30 of 31)

- [Stable Video Diffusion: Scaling Latent Video Diffusion Models to Large Datasets](https://mlsanity.com/paper/2311.15127/) (2023)
- [VBench: Comprehensive Benchmark Suite for Video Generative Models](https://mlsanity.com/paper/2311.17982/) (2023)
- [Towards Accurate Generative Models of Video: A New Metric & Challenges](https://mlsanity.com/paper/1812.01717/) (2018)
- [WorldSimBench: Towards Video Generation Models as World Simulators](https://mlsanity.com/paper/2410.18072/) (2024)
- [DynamiCrafter: Animating Open-domain Images with Video Diffusion Priors](https://mlsanity.com/paper/2310.12190/) (2023)
- [VBench-2.0: Advancing Video Generation Benchmark Suite for Intrinsic Faithfulness](https://mlsanity.com/paper/2503.21755/) (2025)
- [VideoPhy: Evaluating Physical Commonsense for Video Generation](https://mlsanity.com/paper/2406.03520/) (2024)
- [How Far is Video Generation from World Model: A Physical Law Perspective](https://mlsanity.com/paper/2411.02385/) (2024)
- [Towards World Simulator: Crafting Physical Commonsense-Based Benchmark for Video Generation](https://mlsanity.com/paper/2410.05363/) (2024)
- [Do generative video models understand physical principles?](https://mlsanity.com/paper/2501.09038/) (2025)
- [VideoPhy-2: A Challenging Action-Centric Physical Commonsense Evaluation in Video Generation](https://mlsanity.com/paper/2503.06800/) (2025)
- [PHYRE: A New Benchmark for Physical Reasoning](https://mlsanity.com/paper/1908.05656/) (2019)
- [Force Prompting: Video Generation Models Can Learn and Generalize Physics-based Control Signals](https://mlsanity.com/paper/2505.19386/) (2025)
- [VLIPP: Towards Physically Plausible Video Generation with Vision and Language Informed Physical Prior](https://mlsanity.com/paper/2503.23368/) (2025)
- [PISA Experiments: Exploring Physics Post-Training for Video Diffusion Models by Watching Stuff Drop](https://mlsanity.com/paper/2503.09595/) (2025)
- [Good Seed Makes a Good Crop: Discovering Secret Seeds in Text-to-Image Diffusion Models](https://mlsanity.com/paper/2405.14828/) (2024)
- [T2VPhysBench: A First-Principles Benchmark for Physical Consistency in Text-to-Video Generation](https://mlsanity.com/paper/2505.00337/) (2025)
- ["PhyWorldBench": A Comprehensive Evaluation of Physical Realism in Text-to-Video Models](https://mlsanity.com/paper/2507.13428/) (2025)
- [TRAVL: A Recipe for Making Video-Language Models Better Judges of Physics Implausibility](https://mlsanity.com/paper/2510.07550/) (2025)
- [LikePhys: Evaluating Intuitive Physics Understanding in Video Diffusion Models via Likelihood Preference](https://mlsanity.com/paper/2510.11512/) (2025)
- [Simulating the Visual World with Artificial Intelligence: A Roadmap](https://mlsanity.com/paper/2511.08585/) (2025)
- [Phantom: Physics-Infused Video Generation via Joint Modeling of Visual and Latent Physical Dynamics](https://mlsanity.com/paper/2604.08503/) (2026)
- [PhyGround: Benchmarking Physical Reasoning in Generative World Models](https://mlsanity.com/paper/2605.10806/) (2026)
- [PhysChoreo: Physics-Controllable Video Generation with Part-Aware Semantic Grounding](https://mlsanity.com/paper/2511.20562/) (2025)
- [The Invisible Hand of Physics: When Video Diffusion Models Know More Than They Show](https://mlsanity.com/paper/2606.05328/) (2026)
- [Evolution of Video Generative Foundations](https://mlsanity.com/paper/2604.06339/) (2026)
- [Inferring Dynamic Physical Properties from Video Foundation Models](https://mlsanity.com/paper/2510.02311/) (2025)
- [Objects in Generated Videos Are Slower Than They Appear: Models Suffer Sub-Earth Gravity and Don't Know Galileo's Principle...for now](https://mlsanity.com/paper/2512.02016/) (2025)
- [PhyCo: Learning Controllable Physical Priors for Generative Motion](https://mlsanity.com/paper/2604.28169/) (2026)
- [Least-Action-Guided Diffusion for Physical Extrapolation](https://mlsanity.com/paper/2606.11277/) (2026)

## Built on by (most cited first, 2 of 2)

- [How Does Geometry Enter Generated Motion?](https://mlsanity.com/paper/2610.05135/) (2026)
- [World Models' Last Exam in Physics](https://mlsanity.com/paper/2610.08791/) (2026)

## Abstract (arXiv)

Image-to-video generators are often credited with absorbing physical dynamics as implicit world models, a claim the community currently checks with plausibility scores that ask whether a clip looks consistent with real-world motion. Plausibility is the wrong test on its own, because a clip can look natural while encoding the wrong value of the governing physical parameter, and no existing benchmark measures this gap directly. PhysWeep closes it with a fixed, label-free audit, treating a frozen generator as a black box, recovering the realized parameter from generated pixels, and reporting how often generation is trackable at all, how far the realized value sits from the requested one, and which, if either, of the literature's two proposed failure mechanisms the data support. A deterministic-simulator positive control confirms every score is exactly checkable. Applied to three open generators across six sweep axes, PhysWeep finds a specific, reproducible, previously undocumented failure. Conditional on producing trackable motion, two of the three generate confident, well-fit dynamics that converge to one of a small number of fixed, wrong values selected by the sampling seed rather than by the request, reproducing across two independent model families, two physical systems, and an independent tracker. It matches neither the prior reversion nor the case-based clamping the literature anticipates, because the reversion target is seed-conditional rather than a single global default, and a leave-one-out selection rule rejects both; the in-range faithfulness slope is statistically indistinguishable from zero wherever a response is estimable at all. A benchmark averaging over seeds would never see this: each sample is confidently locked to a wrong constant, exactly the failure a plausibility score is structurally blind to. We release the protocol, suite, and analysis code.