# VisPhyWorld: Probing Physical Reasoning via Code-Driven Video Reconstruction

arXiv: https://arxiv.org/abs/2602.13294 · HTML: https://mlsanity.com/paper/2602.13294/

## Key facts

- Cited by 7 papers in the map (data through 14 Oct 2026).
- Placed in Spatial and embodied reasoning, one of 17 directions in LLM reasoning.
- Builds on 47 of its 62 references that are in the map.
- First posted on arXiv 9 Feb 2026.

## Placed in

- LLM reasoning / [Spatial and embodied reasoning](https://mlsanity.com/lr/spatial-and-embodied-reasoning/)
- Multimodal LLMs / vision-language models / [Embodied and 3D vision-language models](https://mlsanity.com/mllm/embodied-and-3d-vision-language-models/)

## Builds on (most cited first, 30 of 47)

- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [The Unreasonable Effectiveness of Deep Features as a Perceptual Metric](https://mlsanity.com/paper/1801.03924/) (2018)
- [Emerging Properties in Self-Supervised Vision Transformers](https://mlsanity.com/paper/2104.14294/) (2021)
- [BERTScore: Evaluating Text Generation with BERT](https://mlsanity.com/paper/1904.09675/) (2019)
- [Stable Video Diffusion: Scaling Latent Video Diffusion Models to Large Datasets](https://mlsanity.com/paper/2311.15127/) (2023)
- [RAFT: Recurrent All-Pairs Field Transforms for Optical Flow](https://mlsanity.com/paper/2003.12039/) (2020)
- [Cosmos World Foundation Model Platform for Physical AI](https://mlsanity.com/paper/2501.03575/) (2025)
- [Image Quality Assessment: Unifying Structure and Texture Similarity](https://mlsanity.com/paper/2004.07728/) (2020)
- [CLEVRER: CoLlision Events for Video REpresentation and Reasoning](https://mlsanity.com/paper/1910.01442/) (2019)
- [Large Language Models Sensitivity to The Order of Options in Multiple-Choice Questions](https://mlsanity.com/paper/2308.11483/) (2023)
- [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)
- [VideoScore: Building Automatic Metrics to Simulate Fine-grained Human Feedback for Video Generation](https://mlsanity.com/paper/2406.15252/) (2024)
- [TheoremQA: A Theorem-driven Question Answering dataset](https://mlsanity.com/paper/2305.12524/) (2023)
- [PhysGen: Rigid-Body Physics-Grounded Image-to-Video Generation](https://mlsanity.com/paper/2409.18964/) (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)
- [Physion: Evaluating Physical Prediction from Vision in Humans and Machines](https://mlsanity.com/paper/2106.08261/) (2021)
- [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)
- [WorldModelBench: Judging Video Generation Models As World Models](https://mlsanity.com/paper/2502.20694/) (2025)
- [StarVector: Generating Scalable Vector Graphics Code from Images and Text](https://mlsanity.com/paper/2312.11556/) (2023)
- [MatPlotAgent: Method and Evaluation for LLM-Based Agentic Scientific Data Visualization](https://mlsanity.com/paper/2402.11453/) (2024)
- [CoPhy: Counterfactual Learning of Physical Dynamics](https://mlsanity.com/paper/1909.12000/) (2019)
- [OmniSVG: A Unified Scalable Vector Graphics Generation Model](https://mlsanity.com/paper/2504.06263/) (2025)
- [MME-Survey: A Comprehensive Survey on Evaluation of Multimodal LLMs](https://mlsanity.com/paper/2411.15296/) (2024)
- [Intuitive physics understanding emerges from self-supervised pretraining on natural videos](https://mlsanity.com/paper/2502.11831/) (2025)
- [GPT4Motion: Scripting Physical Motions in Text-to-Video Generation via Blender-Oriented GPT Planning](https://mlsanity.com/paper/2311.12631/) (2023)
- [Embodied AI Agents: Modeling the World](https://mlsanity.com/paper/2506.22355/) (2025)
- [Physion++: Evaluating Physical Scene Understanding that Requires Online Inference of Different Physical Properties](https://mlsanity.com/paper/2306.15668/) (2023)

## Built on by (most cited first, 7 of 7)

- [Code as Worlds: Agentic Discovery of Executable World Representations for Physical Reasoning](https://mlsanity.com/paper/2608.27549/) (2026)
- [BVB: Benchmarking Agentic Video Understanding via Programmatic Reconstruction in Blender](https://mlsanity.com/paper/2609.15478/) (2026)
- [MPMWorlds: Material-Point-Method Simulations for Inferring and Extrapolating Physical Dynamics](https://mlsanity.com/paper/2606.01538/) (2026)
- [4DCodeBench: Benchmarking Agents on Inverse Graphics of Dynamic Scenes](https://mlsanity.com/paper/2610.03715/) (2026)
- [ExeWorldBench: Benchmarking Code-Based Executable 3D World Modeling](https://mlsanity.com/paper/2610.15166/) (2026)
- [CineRecon: Benchmarking Frontier Models for Agentic Cinematic Reconstruction](https://mlsanity.com/paper/2610.16577/) (2026)
- [PhysCodeGym: Benchmarking Coding Agents for Physically Verifiable 4D Reconstruction](https://mlsanity.com/paper/2610.17288/) (2026)

## Abstract (arXiv)

Evaluating whether Multimodal Large Language Models (MLLMs) genuinely reason about physical dynamics remains challenging. Most existing benchmarks rely on recognition-style protocols such as Visual Question Answering (VQA) and Violation of Expectation (VoE), which can often be answered without committing to an explicit, testable physical hypothesis. We propose VisPhyWorld, an execution-based framework that evaluates physical reasoning by requiring models to generate executable simulator code from visual observations. By producing runnable code, the inferred world representation is directly inspectable, editable, and falsifiable. This separates physical reasoning from rendering. Building on this framework, we introduce VisPhyBench, comprising 209 evaluation scenes derived from 108 physical templates and a systematic protocol that evaluates how well models reconstruct appearance and reproduce physically plausible motion. Our pipeline produces valid reconstructed videos in 97.7% of benchmark runs before fallback. Experiments show that while state-of-the-art MLLMs achieve strong semantic scene understanding, they struggle to accurately infer physical parameters and to simulate consistent physical dynamics. Our code is available https://github.com/TIGER-AI-Lab/VisPhyWorld