Parnassus
Parnassus is a fast detector simulation framework for particle physics. It reads truth-level particle data (HepMC3, ROOT, or Pythia8 input) and produces detector-level outputs using generative models.
Two simulation backends are available:
- Neural: Flow-based generative models that learn detector response from training data
- Parametric: PyTorch-based smearing and efficiency simulation reproducing Delphes-like behavior
Available Generators
| Generator | Type | Description |
|---|---|---|
cms_2011_flow_v00 |
Neural | CMS 2011 flow-based generative model |
aleph_flow_v00 |
Neural | ALEPH flow-based generative model |
cms |
Parametric | CMS detector card (tracker radius 1.29 m, B = 3.8 T) |
atlas |
Parametric | ATLAS detector card (tracker radius 1.15 m, B = 2.0 T) |
aleph |
Parametric | ALEPH detector card (tracker radius 1.5 m, B = 0.435 T) |
Getting Started
- Install Parnassus
- Run your first simulation
- Learn about Neural or Parametric generator modes