Learning & Adaptive Systems Group |

Learning & Adaptive Systems Group

We are part of the Institute for Machine Learning at the Department of Computer Science of ETH Zurich. The group is led by Andreas Krause. Our research is in machine learning and AI, focusing on learning and adaptive systems that actively acquire information, reason and reliably make decisions in complex and uncertain domains. We advance principles of online, active and reinforcement learning, probabilistic and generative modeling for optimization and control, and we engage in real-world applications, typically through interdisciplinary collaboration.

Recent Publications

LITE: Efficiently Estimating Gaussian Probability of Maximality

N. Menet, J. Hübotter, P. Kassraie, A. Krause
In Proc. International Conference on Artificial Intelligence and Statistics (AISTATS), 2025
[bibtex] [abstract] [pdf]

ActSafe: Active Exploration with Safety Constraints for Reinforcement Learning

Y. As*, B. Sukhija*, L. Treven, C. Sferrazza, S. Coros, A. Krause
In Proc. International Conference on Learning Representations (ICLR), 2025
[bibtex] [abstract] [pdf]

MaxInfoRL: Boosting exploration in reinforcement learning through information gain maximization

B. Sukhija, S. Coros, A. Krause, P. Abbeel, C. Sferrazza
In Proc. International Conference on Learning Representations (ICLR), 2025
[bibtex] [abstract] [pdf]

Efficiently Learning at Test-Time: Active Fine-Tuning of LLMs

J. Hübotter, S. Bongni, I. Hakimi, A. Krause
In Proc. International Conference on Learning Representations (ICLR), 2025
Best paper award at Fine-Tuning in Modern Machine Learning Workshop at NeurIPS
[bibtex] [abstract] [pdf]

Stochastic Bilevel Optimization with Lower-Level Contextual Markov Decision Processes

V. Thoma, B. Pásztor, A. Krause, G. Ramponi, Y. Hu
In Proc. Neural Information Processing Systems (NeurIPS), 2024
[bibtex] [abstract] [pdf]

When to Sense and Control? A Time-adaptive Approach for Continuous-Time RL

L. Treven, B. Sukhija, Y. As, F. Dörfler, A. Krause
In Advances in Neural Information Processing Systems (NeurIPS), 2024
[bibtex] [abstract] [pdf]

Progressive Entropic Optimal Transport Solvers

P. Kassraie, A. A. Pooladian, M. Klein, J. Thornton, J. Niles-Weed, M. Cuturi
In Proc. Neural Information Processing Systems (NeurIPS), 2024
[bibtex] [abstract] [pdf]

Bandits with Preference Feedback: A Stackelberg Game Perspective

B. Pásztor, P. Kassraie, A. Krause
In Proc. Neural Information Processing Systems (NeurIPS), 2024
[bibtex] [abstract] [pdf]

News and Highlights

Andreas Krause elected IEEE Fellow.
Charlotte Bunne received an ETH Medal for her Ph.D. thesis.
Parnian Kassraie awarded a Google Ph.D. Fellowship in Machine Intelligence.
Andreas Krause gave an Invited Talk at AutoML 2024.
Charlotte Bunne accepts a faculty position at EPFL.
Giorgia Ramponi accepts a faculty position at University of Zurich.
Andreas Krause elected ACM Fellow.
Andreas Krause appointed to UN High-level Advisory Body on AI.
Charlotte Bunne gave an ICML tutorial on Optimal Transport in Learning, Control, and Dynamical Systems.
Andreas Krause serves as General Chair for ICML 2023.
Ilija Bogunovic accepts a faculty position at UCL.
Andreas Krause awarded Max Planck Fellowship at the Max Planck Institute for Intelligent Systems.
Andreas Krause awarded the Rössler Prize 2021.
Jonas Rothfuss selected as an 2021 Apple Scholar in AI/ML.
Tutorial on Safe Reinforcement Learning at CoRL 2020.
Felix Berkenkamp receives ELLIS Ph.D. Award 2020.
Andreas Krause is co-recipient of ICML Test of Time award 2020.
Hoda Heidari accepts a faculty position at CMU.
Baharan Mirzasoleiman accepts a faculty position at UCLA.
Sebastian Tschiatschek accepts a faculty position at the University of Vienna.
Andreas Krause is co-recipient of ACM SIGKDD Test of Time award 2019.
Kfir Levy accepts a faculty position at the Technion.
Yuxin Chen accepts a faculty position at the University of Chicago.
Andreas Krause gave an Invited Talk on Safe Exploration in Reinforcement Learning at L4DC 2019 at MIT.
New ERC Consolidator project on Reliable Algorithmic Data-Driven Decision Making.
Josip Djolonga received an ETH Medal for his Ph.D. thesis.
Tutorial on Submodularity in Machine Learning: Theory and Applications at Simons Institute 2017.