psetinek/simshift

SIMSHIFT: A Benchmark for Adapting Neural Surrogates to Distribution Shifts

45
/ 100
Emerging

This is a benchmark for researchers and engineers developing AI models to predict outcomes of physical simulations. It helps evaluate how well your 'neural surrogate' models, which are fast AI approximations of complex simulations, perform when the conditions or parameters of the simulation change. You input your AI model and simulation data, and it outputs an evaluation of your model's robustness to these changes, particularly in industrial scenarios like hot rolling or electric motor design.

Use this if you are a researcher or engineer working on AI models for physical simulations and need a standardized way to test your model's performance when simulation parameters or conditions shift.

Not ideal if you are looking for a plug-and-play AI model to run your simulations directly, as this tool is for benchmarking and evaluating such models.

physical-simulation AI-model-evaluation industrial-engineering materials-science mechanical-engineering
No Package No Dependents
Maintenance 10 / 25
Adoption 6 / 25
Maturity 15 / 25
Community 14 / 25

How are scores calculated?

Stars

15

Forks

3

Language

Python

License

MIT

Last pushed

Feb 09, 2026

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/psetinek/simshift"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.