SARL-PLUS

Smart Analytics & Reinforcement Learning

---

Reinforcement Learning and Optimization in Salzburg

---

About Us

Reinforcement Learning and Optimization in Salzburg is a leading organization dedicated to advancing research, education, and innovation in the fields of reinforcement learning and optimization. Our mission is to foster collaboration among researchers, practitioners, and students to explore the latest developments and applications in these areas. We are part of the IDA Lab, which is affiliated with PLUS University Salzburg.

---

Publications

2024

• Hirlaender, S., Kaiser, J., Xu, C., & Santamaria Garcia, A. (2024). Tutorial on Meta-Reinforcement Learning and GP-MPC at the RL4AA'24 Workshop. RL4AA'24 Workshop, Paris-Lodron-Universität Salzburg, Austria, February 5-7, 2024. Zenodo | GitHub

2023

• Hirlaender, S., Lamminger, L., Zevi-Della-Porta, G., & Kain, V. (2023). Ultra Fast Reinforcement Learning in Accelerator Control Demonstrated on CERN AWAKE. In Proceedings of IPAC'23 (pp. 4459-4462). CERN Document Server | InspireHEP

2022

• Eichinger, S., Blanco-García, O., Bruchon, N., Corsini, R., Doebert, S., Farabolini, W., Hirlaender, S., & Latina, A. (2022). Towards Automatic Setup of 18 MeV Electron Beamline Using Machine Learning. Machine Learning: Science and Technology, 4, 025016. DOI: 10.1088/2632-2153/acce21 | arXiv:2209.03183

• Grech, L., Valentino, G., Alves, D., & Hirlaender, S. (2022). Application of Reinforcement Learning in the LHC Tune Feedback. Frontiers in Physics, 10, 929064. DOI: 10.3389/fphy.2022.929064

• Hirländer, S. (2022). Deep Meta Reinforcement Learning for Rapid Adaptation in Linear Markov Decision Processes: Applications to CERN's AWAKE Project. University of Salzburg Repository

2021

• Scheinker, A., Hirlaender, S., Velotti, F. M., Gessner, S., Della Porta, G. Z., Kain, V., & Valentino, G. (2021). Online Multi-objective Particle Accelerator Optimization of the AWAKE Electron Beam Line for Simultaneous Emittance and Orbit Control. Frontiers in Physics, 9, 741639. DOI: 10.3389/fphy.2021.741639

2020

• Kain, V., Hirlander, S., Goddard, B., Velotti, F. M., Della Porta, G. Z., Bruchon, N., & Valentino, G. (2020). Sample-efficient Reinforcement Learning for CERN Accelerator Control. Physical Review Accelerators and Beams, 23(12), 124801. DOI: 10.1103/PhysRevAccelBeams.23.124801 | CERN Document Server

• Hirlaender, S., & Bruchon, N. (2020). Model-free and Bayesian Ensembling Model-based Deep Reinforcement Learning for Particle Accelerator Control Demonstrated on the Fermi FEL. arXiv preprint arXiv:2012.09737. arXiv:2012.09737

• Hirlaender, S., Kain, V., & Schenk, M. (2020). New Paradigms for Tuning Accelerators: Automatic Performance Optimization and First Steps Towards Reinforcement Learning at CERN. In Information, 12(7), 262. DOI: 10.3390/info12070262

---

Contact Information

Team Lead

Simon Hirlaender
Email: [email protected] or [email protected]

University homepage: https://www.plus.ac.at/aihi/der-fachbereich/ida-lab/teams/sarl/

Address

IDA Lab Salzburg
Fachbereich Artificial Intelligence & Human Interfaces (AIHI)
Fakultät Digital & Analytical Sciences (DAS)
Paris Lodron Universität Salzburg (PLUS)
Jakob-Haringer-Straße 6 | Techno 6 | 5020 Salzburg | Austria

SARL Team Website

This is the official website for the Smart Analytics and Reinforcement Learning (SARL) Team at PLUS University Salzburg.

Optimization Notes

The website has been optimized for fast loading with the following techniques:

Performance Optimizations

1. Critical CSS Inlining: Essential styles are inlined in the HTML head to eliminate render-blocking CSS
2. System Font Stack: Using system fonts first reduces font loading time
3. Lazy Loading: Images and non-critical scripts use lazy loading
4. CSS & JS Optimization: Combined and minified files reduce HTTP requests
5. Responsive Design: Optimized for all device sizes

File Structure

---