Ofertes de Feina i Beques Ofertes de Feina
Select Page
Insight Seminar
8 de maig 2025
INSIGHT SEMINAR: Interpret or explain? What to keep in mind when learning physics from neural networks

Hora: Des de 12:00h a 13:00h

Lloc: Elements Room

INSIGHT SEMINAR: Interpret or explain? What to keep in mind when learning physics from neural networks

ANNA DAWID-LEKOWSKA
Leiden University

ABSTRACT:

The importance of machine learning (ML) in quantum physics has been rising, especially in studies of quantum phases of matter or finding ground states of interacting Hamiltonians. The more widespread its use, the more urgent the critical questions of how to extract meaningful physical insights from ML. This talk will explore the differences between explaining a trained model’s behavior (post-hoc explainability) and designing machine learning models with interpretable parts from the ground up. Using two case studies from our research - neural networks applied to the Su-Schrieffer-Heeger (SSH) model and the TetrisCNN model tailored to detecting phase transitions and their order parameters in spin systems - we will show the limitations of post-hoc explainability and the advantages of interpretable architectures. We argue that the most insightful interpretable models are largely task-dependent, and we share our recipe for their design.

BIO:

Anna Dawid is an assistant professor at the Leiden Institute of Advanced Computer Science (LIACS) and the Leiden Institute of Physics (LION) at Leiden University in the Netherlands. Her research interests include interpretable machine learning for science, ultracold quantum simulation platforms, and machine learning theory. Her passion is to transform automated computational methods (especially neural networks) into a new, unique lens, enabling a fresh perspective on difficult, established scientific problems. Before taking up the position in Leiden, she was a researcher at the Center for Computational Quantum Physics at the Flatiron Institute in New York. In 2022, she defended her PhD in physics and photonics under the supervision of Prof. Michał Tomza (Faculty of Physics, University of Warsaw) and Prof. Maciej Lewenstein (ICFO – The Institute of Photonic Sciences, Spain). Previously, she completed her master's degree in quantum chemistry and bachelor's degree in biotechnology at the University of Warsaw. She is the first author of the book "Machine Learning in Quantum Science", which will be published by Cambridge University Press in 2025. She is also a laureate of the START scholarship of the Foundation for Polish Science (2022) and one of the participants of the 74th Nobel Laureate Meeting in Lindau (2024).

Hosted by Prof. Dr. Maciej Lewenstein
Insight Seminar
8 de maig 2025
INSIGHT SEMINAR: Interpret or explain? What to keep in mind when learning physics from neural networks

Hora: Des de 12:00h a 13:00h

Lloc: Elements Room

INSIGHT SEMINAR: Interpret or explain? What to keep in mind when learning physics from neural networks

ANNA DAWID-LEKOWSKA
Leiden University

ABSTRACT:

The importance of machine learning (ML) in quantum physics has been rising, especially in studies of quantum phases of matter or finding ground states of interacting Hamiltonians. The more widespread its use, the more urgent the critical questions of how to extract meaningful physical insights from ML. This talk will explore the differences between explaining a trained model’s behavior (post-hoc explainability) and designing machine learning models with interpretable parts from the ground up. Using two case studies from our research - neural networks applied to the Su-Schrieffer-Heeger (SSH) model and the TetrisCNN model tailored to detecting phase transitions and their order parameters in spin systems - we will show the limitations of post-hoc explainability and the advantages of interpretable architectures. We argue that the most insightful interpretable models are largely task-dependent, and we share our recipe for their design.

BIO:

Anna Dawid is an assistant professor at the Leiden Institute of Advanced Computer Science (LIACS) and the Leiden Institute of Physics (LION) at Leiden University in the Netherlands. Her research interests include interpretable machine learning for science, ultracold quantum simulation platforms, and machine learning theory. Her passion is to transform automated computational methods (especially neural networks) into a new, unique lens, enabling a fresh perspective on difficult, established scientific problems. Before taking up the position in Leiden, she was a researcher at the Center for Computational Quantum Physics at the Flatiron Institute in New York. In 2022, she defended her PhD in physics and photonics under the supervision of Prof. Michał Tomza (Faculty of Physics, University of Warsaw) and Prof. Maciej Lewenstein (ICFO – The Institute of Photonic Sciences, Spain). Previously, she completed her master's degree in quantum chemistry and bachelor's degree in biotechnology at the University of Warsaw. She is the first author of the book "Machine Learning in Quantum Science", which will be published by Cambridge University Press in 2025. She is also a laureate of the START scholarship of the Foundation for Polish Science (2022) and one of the participants of the 74th Nobel Laureate Meeting in Lindau (2024).

Hosted by Prof. Dr. Maciej Lewenstein

Tots els Insight Seminars