Other
geopt
physics-simulation
File size: 2,103 Bytes
8a6eb0e
 
19dbb66
d3e5af5
 
8a6eb0e
 
19dbb66
8a6eb0e
19dbb66
8a6eb0e
19dbb66
 
 
 
 
 
 
 
8a6eb0e
 
 
19dbb66
8a6eb0e
 
 
19dbb66
8a6eb0e
19dbb66
8a6eb0e
 
 
d3e5af5
8a6eb0e
 
 
 
 
 
19dbb66
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
---
license: mit
pipeline_tag: other
tags:
- physics-simulation
---

# GeoPT: Scaling Physics Simulation via Lifted Geometric Pre-Training

[**Project Page**](https://physics-scaling.github.io/GeoPT/) | [**Paper**](https://huggingface.co/papers/2602.20399) | [**GitHub**](https://github.com/Physics-Scaling/GeoPT)

GeoPT is a unified pre-trained model for general physics simulation based on lifted geometric pre-training. It bridges the geometry-physics gap by augmenting geometry with synthetic dynamics, enabling dynamics-aware self-supervision without the need for expensive physics labels.

## Key Features

- **Data Efficiency:** Reduces labeled training data requirements by 20–60% across diverse physics simulation tasks.
- **Scalable Self-Supervision:** Generates millions of training samples quickly, significantly faster than traditional physics supervision.
- **Strong Scaling:** Performance consistently improves with larger models and more training data.
- **Generalization:** Generalizes across diverse physical systems (fluid and solid mechanics) by reconfiguring the dynamics condition as a "prompt".

## Usage

Concrete usage instructions, including environment setup and scripts for fine-tuning GeoPT on various industrial-fidelity benchmarks (such as AirCraft, Cars, and Ships), can be found in the [official GitHub repository](https://github.com/Physics-Scaling/GeoPT).

## Citation

If you find this repo useful, please cite the paper:

```latex
@article{wu2026GeoPT,
  title={GeoPT: Scaling Physics Simulation via Lifted Geometric Pre-Training},
  author={Haixu Wu, Minghao Guo, Zongyi Li, Zhiyang Dou, Mingsheng Long, Kaiming He, Wojciech Matusik},
  booktitle={arXiv preprint arXiv:2602.20399},
  year={2026}
}
```

## Contact

If you have any questions, please contact Haixu Wu (wuhaixu98@gmail.com) and Minghao Guo (guomh2014@gmail.com).

## Acknowledgement

We appreciate the following GitHub repositories for their valuable codebase or datasets: [Transolver](https://github.com/thuml/Transolver) and [Neural-Solver-Library](https://github.com/thuml/Neural-Solver-Library).