physics

As the year is approaching its end, we cordially invite you to this year's last Data Science Colloquium on December 19th, 2024, at 5 pm at the Albert-Einstein-Ring 8-10 on Campus Bahrenfeld (https://g.co/kgs/5eMwmnz, Ground Floor, Rooms 5 & 10). We look forward to talks by DASHH researchers Prof. Christian Schwanenberger and Dr. Andrea Thorn on current opportunities and challenges for artificial intelligence in particle physics & structural biology.

We hope to meet you for discussions after the talks with some snacks & hot beverages in a cheerful atmosphere.

All dates for Event Series "Physics ❤ AI" for the upcoming semester are always at 10:30-12:00 in CFEL SR1.
The detailed agenda is password protected. If you are interested in the agenda, please email the organizers (gregor.kasieczka@uni-hamburg.de / janis.kummer@uni-hamburg.de).
Or just come, it is of course allowed!

All dates for Event Series "Physics ❤ AI" for the upcoming semester are always at 10:30-12:00 in CFEL SR1.
The detailed agenda is password protected. If you are interested in the agenda, please email the organizers (gregor.kasieczka@uni-hamburg.de / janis.kummer@uni-hamburg.de).
Or just come, it is of course allowed!

All dates for Event Series "Physics ❤ AI" for the upcoming semester are always at 10:30-12:00 in CFEL SR1.
The detailed agenda is password protected. If you are interested in the agenda, please email the organizers (gregor.kasieczka@uni-hamburg.de / janis.kummer@uni-hamburg.de).
Or just come, it is of course allowed!

All dates for Event Series "Physics ❤ AI" for the upcoming semester are always at 10:30-12:00 in CFEL SR1.
The detailed agenda is password protected. If you are interested in the agenda, please email the organizers (gregor.kasieczka@uni-hamburg.de / janis.kummer@uni-hamburg.de).
Or just come, it is of course allowed!

In this talk we will propose a model for intelligence inspired by fluid dynamics. By modeling the dynamics of neural activity through a PDE, we achieve a number of interesting properties.

First, we observe that latent representations transform under sensory input transformations as waves, which are in turn well modeled as the evolution of a fluid. We can use this analogy to learn "disentangled representations" of the input, meaning that every input is embedded in a point in latent space which has associated to it a number of tangent directions in which the input transforms in prescribed ways (e.g. rotating, scaling, etc.). We can relate this to equivariant representations but for arbitrary transformations rather than symmetry groups. We show that this strong inductive bias can lead to the formation of topographic organization of neurons, and leads to good predictive performance. We also discuss the reverse, namely how to use deep learning to model PDEs better, in particular the Navier Stokes equations that model the weather. 

Coffee, tea and cookies will be served at 15:30h. After the colloquium we will have beer and pretzels.

The direct link to the webcast is:
https://desy.zoom.us/j/99616528733
Meeting ID: 996 1652 8733
Meeting Password: 733220

The full DESY Particle and Astroparticle Physics Colloquium programme can be found under physikseminar.desy.de/hamburg

A multitude of ML tasks in particle physics, from unfolding detector effects to refining simulation and extrapolating background estimations, require mapping one arbitrary distribution to another. Several indirect methods have been developed to achieve this, such as classifier-based reweighting on a distribution level, or conditional generative models. However, training an ML model to perform a direct, deterministic mapping has long been a challenging prospect.

In this talk, I introduce the concept of Schrödinger Bridges, ML architecture closely related to Diffusion Models, which enables direct mapping of arbitrary distribution to arbitrary distribution. I demonstrate two implementation approaches with differing upsides and present state-of-the-art results applying Schrödinger Bridges to unfolding and refinement tasks.

Institutions

• UHH, Institut für Experimentalphysik

The Network for Artificial Intelligence and Law (NAIL) invites you to its next event. We are pleased to welcome Prof. Dr. Michael Spannowsky to give a talk on the use of neural networks in particle physics. We will learn how neural networks are used to improve our understanding of fundamental scientific questions and how we use mathematical structures observed in nature to better understand neural networks. The lecture will be followed by a discussion around the topic. The event will be held in English.

After the lecture and discussion, we would like to invite you to end the evening with us in a relaxed atmosphere, with pretzels and wine in the south lounge.

You can participate in presence or online. Please register for the event using the following link: Registration

Institutions

• Network for Artificial Intelligence and Law (NAIL)

An exciting talk at the interface of Informatics and Physics by the Rosetta Software Coordinator (European Space Agency), Alessandro Ercolani.

Rosetta is a space mission of the European Space Agency – ESA. It was launched in 2004 and after 10 years of interplanetary travel reached the comet 67P Churyumov-Gerasimenko and sent the lander Philae to its surface: the first time for humankind to touch down on a comet. Alessandro Ercolani was Rosetta Software Coordinator in the Mission Control Team at the time of the launch, and he’ll bring the audience through an exciting trip across our solar system with many pictures and animations, to tell the story of this fantastic exploration mission.

Institutions

• UHH