News Overview

We are happy to announce the next Scientific Ma­chine Learning event from the series “Ma­chine Learning Galore!”, taking place on April 25, 4:30 to 6:00 pm at INF 205 Mathematikon (5th floor). The event will include various lab presentations and science talks:

Ma­chine Learning galore! - Programme:

• Lab presentations:
  • Tobias Buck
  • Holger Fröning
  • Denis Schapiro
• Rocket Science:
  • William Oliver (Buck lab): Ma­chine Learning for cosmological simulations
  • Hendrik Borras (Fröning lab): Probabilistic Photonic Computing with Chaotic Light
  • Miguel A. Ibarra Arellano (Schapiro lab): Finding broken cells with AI and computer vision

In order to participate, please register for free at https://www.mlai.uni-heidelberg.de/en/machine-learning-talks-on-campus.

Scientific Ma­chine Learning is a joint initiative from IWR and STRUC­TURES to foster interactions within and development of the local ma­chine learning community. Its portal, http://mlai.uni-heidelberg.de summarizes the many relevant events and news from across campus that would otherwise remain scattered across single institutions or fields. The goal of the initiative aligns with the STRUC­TURES Cluster of Excellence's objective of driving re­search into the fundamental understanding of current and future ma­chine learning, and with IWR’s aim to leverage ma­chine learning to enable the solution of long-standing problems in the natural and life sciences, the engineering sciences, as well as the humanities.

Further information:

• Scientific Ma­chine Learning Website
• Ma­chine Learning Talks on Campus – Information service and mailing list
• Interdisciplinary Center for Scientific Computing (IWR)

Simulations provide a potential explanation for the mysterious gap in the size distribution of super-Earths.

Astronomers from Germany and Switzerland, including STRUC­TURES member Thomas Henning (Director of Max Planck Institute for Astronomy, MPIA) and YRC Postdoc Remo Burn (MPIA), have uncovered evidence of how the enigmatic gap in the size distribution of exoplanets at around two Earth radii emerges – a longstanding problem in exoplanetary research. Their computer simulations demonstrate that the migration of icy, so-called sub-Neptunes into the inner regions of their planetary systems could account for this phenomenon. Ordinarily, planets in evolved planetary systems, such as the Solar System, follow stable orbits around their central star. However, many indications suggest that some planets might depart from their birthplaces during their early evolution by migrating inward or outward. As they draw closer to the central star, evaporating water ice forms an atmosphere that makes the planets appear larger than in their frozen state. Simultaneously, smaller rocky planets gradually lose a portion of their original gaseous envelope, causing their measured radius to shrink over time. The authors reported their findings in an article published in Nature Astronomy. (Original Press Release by MPIA)

Further information:

• Press release by Max Planck Institute for Astronomy (MPIA) with more details
• Original publication in Nature Astronomy: Remo Burn, Christoph Mordasini et al. (2024): A radius valley between migrated steam worlds and evaporated rocky cores

The latest issue of Ruperto Carola, Hei­del­berg Uni­ver­si­ty's re­search magazine, explores the pressing question of artificial intelligence's impact on society and the broader themes of “Right & Wrong.” From historical deceptions to modern dilemmas like fake news and conspiracy theories, this issue presents insightful perspectives contributed by a range of researchers, including STRUC­TURES members Georgia Koppe, Manfred Salmhofer, and Fred Hamprecht.

In a collaborative article, computational psychiatrist Georgia Koppe from STRUC­TURESalongside her colleagues Martin Fungisai Gerchen (Biological Psychology) and Andreas Meyer-Lindenberg (Psychiatry and Psychotherapy), intricately examine the interplay between various forms of heterodox beliefs, with a focus on conspiracy theories. They shed light on their characteristics, prevalence, and potential correlations with mental health factors. In the same issue, Fred Hamprecht from STRUC­TURES (Ma­chine Learning and Artificial Intelligence) engages in a conversation with ChatGPT on AI ethics. In the interview “Proofs of God and Flight Paths – The Search for Universal Truths,” mathematical physicist and speaker of the STRUC­TURES Cluster, Manfred Salmhofer and theologian Prof. Dr. Friederike Nüssel discuss the subjects of knowledge creation and the concept of “right”, “wrong” and universal truths in science.

The Rupert Carola re­search magazine appears twice a year, mainly in German with English abstracts, and addresses all members of the uni­ver­si­ty, its partners in academia, politics, business and society, and, in particular, alumni, friends and an interested public in Germany and abroad. All editions are accessible via the open access publisher Hei­del­berg Uni­ver­si­ty Publishing (heiUP).

Further information:

• Ruperto Carola Issue 23 · March 2024
• Georgia Koppe's Website
• Manfred Salmhofer's Website
• Fred Hamprecht's Website

We are happy to present the 15th volume of the STRUC­TURES Newsletter, featuring insightful news, background articles and interviews. This edition features the following topics:

1. Probing Universal Dynamics with Topological Data Analysis in a Gluonic Plasma
2. EP 5.4: The Quest for an Unknown Functional That We Know Exists
3. STRUC­TURES Welcomes its First YAM Fellows
4. STRUC­TURES Asks: Freya Jensen and Victoria Noel (EP Math & Data)

Additionally, we invite you to engage with the interactive version of our newsletter's cover image, containing a visualization by Tobias Kaczun & Roman Remme.

The STRUC­TURES Project Management Office is happy to answer questions and to receive feedback.

The STRUC­TURES Cluster, in collaboration with the 4EU+ Eu­ro­pean Uni­ver­si­ty Alliance, is thrilled to announce the spring block course Quan­tum Information and Quan­tum Many-Body Theory taking place from March to June 2024. The course includes both online lectures and in-person events at Uni­ver­si­ty of Copenhagen. The course is aimed at Master and PhD students, yet it warmly welcomes all interested students, extending an invitation even to those not affiliated with a uni­ver­si­ty that is part of the 4EU+ Alliance.

Title: Quan­tum Information and Quan­tum Many-Body Theory
Date and Location:

• Phase 1: 7 March - 2 May 2024 (Online)
• Phase 2: 2-31 May 2024 (Poster session)
• Phase 3: 24-28 June 2024 (Physical Master Class)

Course Description: In an era marked by the second quan­tum revolution and the rise of quan­tum technological advancements, the necessity for a deep understanding of quan­tum many-body systems and quan­tum information theory has never been more critical. Our course aims to equip participants with comprehensive knowledge and proficiency in these areas, focusing on the mathematics underpinning quan­tum theory, the characterization of complex quan­tum systems, and the development of algorithms for quan­tum information processing.
Travel Grants: Available for 4EU+ Alliance students
Registration Deadline: Registration is possible until 1 March 2024.

The goal is to provide students with a blended learning approach to the course content on the mathematics of quan­tum theory with special emphasis on complex quan­tum many-body systems and quan­tum information theory. The main intended learning outcomes include an overview of the latest developments in those fields. In addition, the course aims to help the participants to develop competencies to engage in self-organized cross-university and interdisciplinary collaborations via online team work as well as to give and receive feedback on results.

The 4EU+ Eu­ro­pean Uni­ver­si­ty Alliance is a transnational strategic association bringing together seven comprehensive, research-intensive, public universities from four regions of Europe: Charles Uni­ver­si­ty in Prague, Hei­del­berg Uni­ver­si­ty, Paris-Panthéon-Assas Uni­ver­si­ty, Sorbonne Uni­ver­si­ty in Paris, the Uni­ver­si­ty of Copenhagen, the Uni­ver­si­ty of Geneva, the Uni­ver­si­ty of Milan, and the Uni­ver­si­ty of Warsaw. The 4EU+ Alliance has one vision: to create one comprehensive research-intensive Eu­ro­pean Uni­ver­si­ty through a new quality of cooperation in teaching, education, re­search and administration.

Further information:

• Course Website (University Copenhagen) with further details
• 4EU+ Eu­ro­pean Uni­ver­si­ty Alliance

We are delighted to announce the upcoming talk by Prof. Klaus-Robert Müller at the IWR Colloquium on Tuesday, Jan 30, 2024. Prof. Klaus-Robert Müller, a renowned authority in the German ma­chine learning community, will be sharing his expertise on Ma­chine Learning and AI for the sciences, with a particular focus on chemistry and physics:

Title: Ma­chine Learning and AI for the Sciences: Toward Understanding
Speaker: Prof. Klaus-Robert Müller (Director, Berlin Institute for the Foundations of Learning & Data · ML Group, TU Berlin)
Abstract: In recent years, ma­chine learning (ML) and artificial intelligence (AI) methods have begun to play a more and more enabling role in the sciences and in industry. In particular, the advent of large and/or complex data corpora has given rise to new technological challenges and possibilities. In his talk, Müller will touch upon the topic of ML applications in the sciences, in particular in chemistry and physics. He will also discuss possibilities for extracting information from ma­chine learning models to further our understanding by explaining nonlinear ML models. Finally, Müller will briefly discuss perspectives and limitations.
Date/Time: Tuesday, Jan 30, 2024, at 16:15
Location: Mathematikon, Conference Room / 5th Floor, INF 205, 69120 Hei­del­berg (and online via Zoom.)

Prior to the IWR colloquium, there will be a get-together for all members of the HGS MathComp at 15:30 at the common room / fifth floor.

Further information:

• Talk announcement poster

We are excited to announce the inaugural Chemical Compound Space Conference (CCSC2024), which will take place on May 21-24, 2024 in Hei­del­berg. The conference aims to bring together the vibrant community of chemists, physicists, and data scientists who employ quan­tum ma­chine learning to gain physics-informed understanding of the vast chemical compound space. Engaging talks by renowned experts, stimulating discussions, and valuable networking opportunities await the participants. We extend a cordial invitation to everyone interested – whether experts, early- and mid-career academics, postdocs, and students – for this enriching and collaborative event.

Further information:

• Conference website with further details
• Registration portal
• Abstract submission: You can apply for a short talk, poster, or digital demo by submitting your abstract via an abstract submission portal. A template is provided in the portal. The deadline is 15.03.2024.