Speaker
Description
The increasing demands on simulation statistics for HL-LHC analyses challenge the scalability of traditional calorimeter simulation within all LHC collaborations. Fast simulation techniques based on machine learning have proven effective, yet further improvements may arise from quantum-inspired models.
We investigate the feasibility of integrating Quantum Neural Network (QNN) components into the ATLAS fast calorimeter simulation framework, focusing on two complementary approaches. The first one employs a hybrid quantum–classical pipeline where a QNN is used to learn and generate the latent space of calorimeter shower representations, subsequently integrated into a classical Generative Adversarial Network (GAN) for sample generation. The second approach explores a quantum-inspired generative model based on Invertible Neural Networks (INN), providing a reversible mapping between input kinematic variables and calorimeter observables, thereby enabling explicit likelihood evaluation and enhancing interpretability. We report on implementation details, performance and ideas for future development.
| Sessions | Quantum Machine Learning: |
|---|---|
| Invited | No |