Publications of relAI students at ICLR 2025

The excellent work of relAI students will be prominently represented at the Thirteenth International Conference on Learning Representations (ICLR) 2025, which will take place at the Singapore EXPO from 24 to 28 April 28 2025.

Twelve publications from our students will be presented at the conference, nine of them in the main track. Notably, four out of these nine publications have been selected for Oral or Spotlight presentations. This is a significant achievement and demonstrates the high quality of relAI research, considering that only 15% of accepted papers are invited to give a talk.

If you plan to attend the conference, do not miss the opportunity to discuss these publications directly with some of our students. Be sure to attend the Oral presentation by Yan Scholten titled "A Probabilistic Perspective on Unlearning and Alignment for Large Language Models" on the 24th April. You can learn about Lisa Wimmer´s work, "Trust Me, I Know the Way: Predictive Uncertainty in the Presence of Shortcut Learning" at the Workshop on Spurious Correlation and Shortcut Learning: Foundations and Solutions. Additionally, check out the posters of Amine Ketata and Chengzhi Hu!.  Amine will be presenting his work on “Lift Your Molecules: Molecular Graph Generation in Latent Euclidean Space” and you can talk to Chengzhi Hu about “Surgical, Cheap, and Flexible: Mitigating False Refusal in Language Models via Single Vector Ablation”.

Full list of relAI publications at ICLR 2025:

    Oral Presentation - Main Track

  1. A Probabilistic Perspective on Unlearning and Alignment for Large Language Models
    Yan Scholten, Stephan Günnemann, Leo Schwinn
  2. Spotlight Presentation - Main Track

  3. Exact Certification of (Graph) Neural Networks Against Label Poisoning
    Mahalakshmi Sabanayagam, Lukas Gosch, Stephan Günnemann, Debarghya Ghoshdastidar
  4. Provably Reliable Conformal Prediction Sets in the Presence of Data Poisoning
    Yan Scholten, Stephan Günnemann
  5. Signature Kernel Conditional Independence Tests in Causal Discovery for Stochastic Processes
    Georg Manten, Cecilia Casolo, Emilio Ferrucci, Søren Wengel Mogensen, Cristopher Salvi, Niki Kilbertus
  6. Posters - Main Track

  7. Differentially private learners for heterogeneous treatment effects
    Maresa Schröder, Valentyn Melnychuk, Stefan Feuerriegel
  8. Surgical, Cheap, and Flexible: Mitigating False Refusal in Language Models via Single Vector Ablation
    Xinpeng Wang, Chengzhi Hu, Paul Röttger, Barbara Plank
  9. ParFam -- (Neural Guided) Symbolic Regression via Continuous Global Optimization
    Philipp Scholl, Katharina Bieker, Hillary Hauger, Gitta Kutyniok
  10. Lift Your Molecules: Molecular Graph Generation in Latent Euclidean Space
    Mohamed Amine Ketata, Nicholas Gao, Johanna Sommer, Tom Wollschläger, Stephan Günnemann
  11. Constructing Confidence Intervals for Average Treatment Effects from Multiple Datasets
    Yuxin Wang, Maresa Schröder, Dennis Frauen, Jonas Schweisthal, Konstantin Hess, Stefan Feuerriegel
  12. Workshops

  13. Trust Me, I Know the Way: Predictive Uncertainty in the Presence of Shortcut Learning
    Lisa Wimmer, Bernd Bischl, Ludwig Bothmann
    Workshop on Spurious Correlation and Shortcut Learning: Foundations and Solutions
  14. Privacy Amplification by Structured Subsampling for Deep Differentially Private Time Series Forecasting
    Jan Schuchardt, Mina Dalirrooyfard, Jed Guzelkabaagac, Anderson Schneider, Yuriy Nevmyvaka, Stephan Günnemann
    Workshop on Advances in Financial AI: Opportunities, Innovations and Responsible AI
  15. Cracking the Code: Evaluating Zero-Shot Prompting Methods for Providing Programming Feedback
    Niklas Ippisch, Anna-Carolina Haensch, Markus Herklotz, Jan Simson, Jacob Beck, Malte Schierholz
    Workshop on Human-AI Coevolution
  16. Exact Certification of (Graph) Neural Networks Against Label Poisoning
    Mahalakshmi Sabanayagam, Lukas Gosch, Stephan Günnemann, Debarghya Ghoshdastidar
    VerifAI: AI Verification in the Wild

🥂Congratulations!

We are proud to announce that our relAI Director Gitta Kutyniok has been invited to become a member of the US National Academy of Artificial Intelligence (NAAI). NAAI is committed to advancing artificial intelligence by fostering collaboration among leading experts and promoting innovative research and development.

The election acknowledges Gitta Kutyniok's distinguished contributions to applied harmonic analysis, compressed sensing, and artificial intelligence. This honor recognizes her leadership as the Bavarian AI Chair for Mathematical Foundations of Artificial Intelligence at Ludwig-Maximilians-Universität München (LMU Munich), which has significantly advanced research and collaboration in these fields. Additionally, NAAI greatly appreciated her recognition as a SIAM Fellow in 2019 and an IEEE Fellow in 2024, which highlights her outstanding accomplishments and the high regard in which she is held by her peers. 

relAI is proud to support the opening ceremony of the project “Next Generation AI Computing (gAIn)"! This collaboration between LMU and TUM in Bavaria and TU Dresden in Saxony is financially supported by the Bavarian State Ministry of Science and the Arts, as well as the Saxon State Ministry for Science, Culture, and Tourism. The goal of the project is to develop a comprehensive, mathematics-based concept for the next generation of "Green AI" systems. The focus will be on application-based AI hardware-software combinations aimed at maximizing energy efficiency, trustworthiness, and legal compliance.

The event will feature keynotes from Minister Blume (Bavaria) and Minister Gemkow (Saxony), and a scientific lecture by Prof. Dr. Wolfram Burgard (UTN), as well as general information about the project.

For the Konrad Zuse Schools of Excellence in Artificial Intelligence, the project presents a significant opportunity to collaborate, as relAI Director Gitta Kutyniok and Stefanie Speidel, deputy director of the Zuse School SECAI, are both Principal Investigators on the project.

 

🎉The 8th edition of DataFest Germany will be held at Ludwig-Maximilians-Universität in Munich from 28 March to 30 March 2025. relAI is proud to support the event organization again this year. Additionally, a team of relAI students will participate in this exciting competition and networking opportunity.

The event is an annual data-driven competition, commonly referred to as a “hackathon,” that alternates between Mannheim and Munich. It is organized in collaboration with partners from industry and research institutions.

Datafest Germany is a celebration that follows upon the model DataFest™, organized by the American Statistical Association. The world's first DataFest took place at the University of California in 2011. Since then, many universities took up the DataFest format.

Learn more about DataFest Germany in this link.

Save the date!

We proudly invite you to our next Munich AI Lecture. This is the flagship speaker series about AI in Munich, co-organized by relAI.

 Event Details:

  • Speaker: Prof. Michael Mahoney (UC Berkeley)
  • Title: Foundational Methods for Foundation Models for Scientific Machine Learning
  • Date and Time: March 26, 2025 14:00 CET
  • Location: Lecture Hall W201, Professor-Huber-Platz 2, LMU Munich, 80539 Munich (Metro U3/U6 Universität, Exit B) LMU Room Finder

Abstract

The remarkable successes of ChatGPT in natural language processing (NLP) and related developments in computer vision (CV) motivate the question of what foundation models would look like and what new advances they would enable, when built on the rich, diverse, multimodal data that are available from large-scale experimental and simulational data in scientific computing (SC), broadly defined. Such models could provide a robust and principled foundation for scientific machine learning (SciML), going well beyond simply using ML tools developed for internet and social media applications to help solve future scientific problems. Prof. Mahoney will describe recent work demonstrating the potential of the "pre-train and fine-tune" paradigm, widely-used in CV and NLP, for SciML problems, demonstrating a clear path towards building SciML foundation models; as well as recent work highlighting multiple "failure modes" that arise when trying to interface data-driven ML methodologies with domain-driven SC methodologies, demonstrating clear obstacles to traversing that path successfully. Prof. Mahoney will also describe initial work on developing novel methods to address several of these challenges, as well as their implementations at scale, a general solution to which will be needed to build robust and reliable SciML models consisting of millions or billions or trillions of parameters.

Bio of the speaker

Michael W. Mahoney is at the University of California at Berkeley in the Department of Statistics and at the International Computer Science Institute (ICSI). He is also an Amazon Scholar as well as head of the Machine Learning and Analytics Group at the Lawrence Berkeley National Laboratory. He works on algorithmic and statistical aspects of modern large-scale data analysis. Much of his recent research has focused on large-scale machine learning, including randomized matrix algorithms and randomized numerical linear algebra, scientific machine learning, scalable stochastic optimization, geometric network analysis tools for structure extraction in large informatics graphs, scalable implicit regularization methods, computational methods for neural network analysis, physics informed machine learning, and applications in genetics, astronomy, medical imaging, social network analysis, and internet data analysis. He received his PhD from Yale University with a dissertation in computational statistical mechanics, and he has worked and taught at Yale University in the mathematics department, at Yahoo Research, and at Stanford University in the mathematics department. Among other things, he was on the national advisory committee of the Statistical and Applied Mathematical Sciences Institute (SAMSI), he was on the National Research Council's Committee on the Analysis of Massive Data, he co-organized the Simons Institute's fall 2013 and 2018 programs on the foundations of data science, he ran the Park City Mathematics Institute's 2016 PCMI Summer Session on The Mathematics of Data, he ran the biennial MMDS Workshops on Algorithms for Modern Massive Data Sets, and he was the Director of the NSF/TRIPODS-funded FODA (Foundations of Data Analysis) Institute at UC Berkeley. More information is available at https://www.stat.berkeley.edu/~mmahoney/ .

This event is open to everyone; registration is not required.

The Roland Berger Foundation (RBS) and TUM have begun a collaboration to promote the AI skills of socially disadvantaged children and young people. RBS works with 70 partner schools throughout Germany to provide scholarships to talented primary school pupils from the second grade onwards from socially disadvantaged families.

relAI Fellow Enkeledja Kasneci is the scientific director of the project. The scholarship holders learn how to use AI responsibly and reflectively. AI tools are also being developed to better support children and young people with difficult starting conditions

For more information, please visit the websites of Roland Berger Foundation and TUM.

The European legal initiative to regulate AI (Artificial Intelligence Act) poses a particular challenge to small and medium-sized enterprises (SMEs) and start-ups that want to benefit from artificial intelligence and pursue innovations. Bavarian AI Act Accelerator, a new project funded by the Bayerisches Staatsministerium für Digitales and coordinated by the appliedAI Institute for Europe gGmbH, is designed to support companies in fulfilling the new requirements and, therefore, lower barriers to the use of artificial intelligence.

Principal contributors of the project include relAI directors Prof. Dr.  Gitta Kutyniok and Prof. Dr. Stephan Günnemann,  relAI fellow Prof. Dr. Mark Zöller, as well as scientists from the Technical University of Munich (TUM) and the University of Technology Nuremberg (UTN), who provide the necessarily high degree of interdisciplinarity.

relAI director Prof. Dr. Gitta Kutyniok leads the scientific part of the project. One of the main goals is to develop a system for automatic, and hence easy and fair, verification with the EU AI Act. This requires the following steps:
🔹Derive a profound legal understanding of the different terminologies.
🔹Develop a formalization/mathematization of the articles.
🔹 Build a system for automatic verification.

🛫 Our director, Gitta Kutyniok, gave a talk and joined the panel discussion at the Kick-Off event last week (photo) as the scientific lead of the project.   

Excerpts taken from:

After Stargate and DeepSeek, which technological developments will influence the future of the AI race? What implications does this hold for Germany and Europe? Have we constrained ourselves too soon with the AI Act? In an interview on the Plattform Lernende Systeme website, Prof. Dr. Gitta Kutyniok, Director of relAI and member of the Platform, discusses the current dynamics and explains how mathematics can enhance the comprehensibility of AI results.

Watch the interview here.

The Saxon-Bavarian AI project GAIn – Next Generation AI Computing is a pilot project tackling new AI hardware and software concepts to reduce energy consumption and increase reliability for different applications such as surgical robotics. It builds on the foundation of the Cluster of Excellence CeTI, the 6G-life research hub, and the Konrad Zuse Schools of Excellence SECAI and relAI. The project aims to address key challenges in energy consumption, predictability, reliability, and legal implementation. A core objective is to significantly reduce the energy consumption of AI-based applications while enhancing their predictability and reliability for different applications such as surgical robotics.

The project has now been officially launched. Together with Frank Fitzek (TU Dresden), Gitta Kutyniok (LMU, relAI) and Holger Boche (TUM), Stefanie Speidel (TU Dresden, SECAI) hosted the kick-off meeting of the project.  The cooperation across federal states will strengthen Germany's technological sovereignty and contribute to the international leadership role of Saxony and Bavaria in central computing technologies.

Excerpts from the TUD press release AI project "GAIn" with TUD participation aims to propel Saxony and Bavaria to an international leadership role in computing technologies, of the National Center of Tumor Disease Dresden (NCT)  GAIn (2024 – 2027)  and of SECAI news https://secai.org/content/news/56

relAI warmly welcomes LMU Professor David Rügamer to our school. David heads the Data Science Group at LMU, and he is also a Principal Investigator at the Munich Center for Machine Learning (MCML).

Prof. Rügamer works on fundamental topics within the relAI research area Mathematical and Algorithmic Foundations applied to neural networks, such as symmetries, sparsity, and uncertainty quantification in deep neural networks. Additionally, his work is also relevant to the Algorithmic Decision-Making relAI research topic. relAI will benefit from his research experience and, furthermore, from his contributions to our Curriculum, including lectures..