Multi-canonical methods and lattice field theory

Aula Renzo Leonardi - Villa Tambosi

Str. delle Tabarelle, 286, 38123 Villazzano TN
Villazzano

***WARNING: for hotel bookings do not give your data to any external services upon email request; there are known cases of fraud***

 

This workshop will focus on lattice field theory (LFT) methods that simulate nuclear and particle physics systems simultaneously across a range of temperatures. These "multi-canonical" methods provide a natural way to study thermodynamic phases, phase transitions, and criticality in many systems. In contrast to mainstream methods for LFT, which require enormous computational power to achieve robust results, multi-canonical methods have been demonstrated to avoid algorithmic issues that obstruct efficient computational scaling. Hence these methods provide a promising approach for future lattice calculations, especially in the direction of highly-challenging LFT calculations in Quantum Chromodynamics. However, input from other research fields will be crucial in developing state-of-the-art techniques, and the workshop is broadly scoped to include experts in these methods from other domains, such as condensed matter, molecular systems, and Bayesian parameter estimation.
The techniques to be discussed span two general classes:
1. Energy-based methods: density-of-states and nested sampling
2. Temperature-based methods: parallel tempering and out-of-equilibrium simulation
Research on multi-canonical methods has also recently gained momentum due to the benefits achieved by connecting the above methods with machine learning (ML) techniques, including flow models and diffusion models. The discourse during the workshop will cover the development of these classes of methods with the aim of answering the following central questions:
* What are the common patterns in these approaches from which shared expertise can be developed?
* What new directions should be pursued, especially in connection with rapid developments of related ML techniques?
Emphasis will be placed on research directions that accelerate the sampling algorithms required across all LFT calculations and on those enabled by new ML results.
The workshop will be open to participants with backgrounds in LFT and in other domains, as well as those with experience in related ML methods. This scope of participants is selected to encourage exchanges between practitioners of multi-canonical methods within and outside LFT and to spark ideas on further incorporating ML methods. The multidisciplinary background of the key speakers confirm that our proposal is timely and has the potential to attract further leading experts across multiple fields.
It is an important goal of this workshop to advance early-career researchers such that they can become leaders in this rapidly developing research area. This has been reflected in the key invited speakers and in the organizing committee (five of the members are early-career researchers). Speakers have also been invited with a consideration for diversity, in particular including gender diversity.

Organizers

  • Gurtej Kanwar (University of Bern)
  • Claudio Bonanno (IFT UAM/CSIC Madrid)
  • Daniel Hackett (Fermilab)
  • Biagio Lucini (Swansea University)
  • Julian Urban (MIT)
See complete details and information

Registration

Registration available from 29/09/2025 until 07/11/2025.

Privacy Notice

Pursuant to art. 13 of EU Regulation No. 2016/679 – General Data Protection Regulation and as detailed in the Privacy Policy for FBK-ECT* event’s participants, we inform you that the event will be recorded and disclosed on the FBK-ECT* institutional channels. In order not to be filmed or recorded, you can disable the webcam and/or mute the microphone during virtual events or inform the FBK-ECT* staff who organize the public event beforehand.
WordPress Cookie Notice by Real Cookie Banner