Announcements

Upcoming events


  • Postponed! Bethe Forum "Inflation", March 23 - 27, 2020
    Organizers: M. Peloso (Padua), M. Drees (Bonn), H. P. Nilles (Bonn) and Ivonne Zavala (Swansea)
  • Postponed! Bethe-Colloquium by Martin Reuter (JGU Mainz), April 23rd, 2020
  • Postponed! Bethe Forum "Modular Flavor Symmetries", May 4-8, 2020
    Organizers: M. Ratz (Irvine), F. Feruglio (Padua), T. Kobayashi (Sapporo) and H. P. Nilles (Bonn).
  • Postponed! Bethe Forum "Machine Learning: Where to Apply in Theoretical Physics", June 1-5, 2020
    Organizers: F. Ruehle (CERN), J. Halverson (Boston), H. P. Nilles (Bonn) and P. Vaudrevange (Munich).
  • Bethe Colloquium by Fabian Ruehle (CERN), June 4th, 2020
    This Colloquium will be held as a ZOOM meeting.
  • Bethe Colloquium by Ferruccio Feruglio (University of Padova), June 18th, 2020
    This Colloquium will be held as a ZOOM meeting.
  • Bethe Colloquium by Richard Dawid (Stockholm University): postponed from June 25th to November 12th, 2020

Further information will be given as soon as it is available.

Bethe Forum "Machine Learning: Where to Apply in Theoretical Physics"

01.06.2020 – 05.06.2020
Poster Forum Machine Learning
Postponed! - Due to the Covid-19 situation, the Bethe Forum "Modular Flavor Symmetry" is postponed. The future date will be announced in time.





Bethe Colloquium by Fabian Ruehle

June 2020

The next Bethe Colloquium is scheduled for Thursday, June 4th, at 4:15 pm, via ZOOM.

  • Fabian Ruehle (CERN, Geneva)
  • Machine Learning in Theoretical Physics
  • ZOOM-meeting

Abstract: Machine Learning has become an integrable part of our daily lives. Owing to its versatile and universal nature, it is also increasingly utilized in Science. I will first provide an overview of different machine learning techniques and how they are applied in Physics. After that, I will discuss example applications from my own field of research. In a first application, we use deep neural networks to solve the differential geometry problem of finding (complex Kahler) metrics with vanishing Ricci tensor, leading to so-called Calabi-Yau metrics. In a second example, we use Reinforcement Learning to solve the combinatorial problem of finding integer solutions to a coupled system of polynomial equations. We demonstrate that the Reinforcement Learning algorithm discovers human-derived strategies, but also finds new, more efficient ones. Lastly, we adopt a recent development from the field of natural language processing, called the Reformer Neural Network, to study knot theory.

If you are interested in attending the ZOOM-Meeting, please send an email to theory@physik.uni-bonn.de. We will send the access details shortly before the Colloquium.

Bethe Colloquium by Ferruccio Feruglio

June 2020

A second Bethe Colloquium in June is scheduled for Thursday, 18th, at 4:15 pm, via ZOOM.

  • Ferruccio Feruglio (University of Padova)
  • Modular Flavour Symmetries
  • ZOOM-meeting

Abstract: The hope of relating fermion masses and mixing angles to some fundamental underlying principle has lead the scientific community to explore a vast variety of possible symmetry patterns and their origin. Such effort, which will be briefly reviewed in this talk, has revealed that realistic candidate symmetries should be substantially broken, thus raising the problem of a correct description and interpretation of the symmetry breaking sector. In this talk I will discuss an approach to the flavour puzzle where the role of flavour symmetry is played by modular invariance, ubiquitous in string theory and in condense matter systems. I will explain why modular invariance is, in a sense, always a broken symmetry and how we can naturally interprete its symmetry breaking sector as arising from the the simplest compactifications of extra dimensions, as realized in string theory. I will describe some models of neutrino masses and mixing angles, stressing the remarkable and unique properties of modular invariance in this context..

If you are interested in attending the ZOOM-Meeting, please send an email to theory@physik.uni-bonn.de. We will send the access details shortly before the Colloquium.

Further events will be announced as soon as possible.