Department of Physics, National University of Singapore
Condensed Matter Seminar Series
To receive email notifications of this seminar series, please send an email to zouliujun@gmail.com to request.
Condensed Matter Seminar Series
To receive email notifications of this seminar series, please send an email to zouliujun@gmail.com to request.
Upcoming Seminars
(Both online and in-person seminars will be broadcast via Zoom.)
(Click the upper right corner of this page to see the previous seminars and their recordings.)
Time: 9am, October 23 2026, GMT+8
Location: Online
Zoom: https://nus-sg.zoom.us/j/88911975409?pwd=ixb7zXPuROkgclFcXZ2Wzh5vx6fmz4.1
Password: 835700
Speaker: Ruizhi Liu
Title: Haag duality from conditional mutual information
Abstract: Haag duality is a widely used assumption in quantum many-body physics, particularly in the study of anyons and topological order, yet establishing it beyond commuting projector models remains challenging. In this talk, I will present an entanglement-based characterization of Haag duality: it holds if and only if a suitable conditional mutual information vanishes in an appropriate limit. Applicable to arbitrary spatial regions in any dimension, this criterion involves only entanglement entropies of finite subsystems and provides a relaxed form of the entanglement bootstrap axiom A1. I will discuss its implications for topological entanglement entropy and Lieb–Schultz–Mattis-type constraints, including the equivalence between half-chain Haag duality and vanishing entropy density in one dimension.
Based on joint work with Lauritz van Luijk ([arXiv:2609.13550](https://arxiv.org/abs/2609.13550)).
Time: 9am, October 30 2026, GMT+8
Location: Online
Zoom: https://nus-sg.zoom.us/j/88911975409?pwd=ixb7zXPuROkgclFcXZ2Wzh5vx6fmz4.1
Password: 835700
Speaker: Jacob Zavatone-Veth
Title: The many lives of linear regression
Abstract: Though linear regression is one of the most thoroughly studied problems in statistical learning, it continues to surprise us. In this talk, I will explore how linear regression provides a common framework for understanding a wide range of high-dimensional learning problems, from classical questions about generalization from limited data to contemporary developments in generative diffusion models and in-context learning. Along the way, I will highlight the common mathematical structures that enable us to gain detailed insight. I will conclude with some general thoughts about how progress in understanding linear regression informs what we might ask of a broader physics of learning.
Time: 9am, November 6 2026, GMT+8
Location: Online
Zoom: https://nus-sg.zoom.us/j/88911975409?pwd=ixb7zXPuROkgclFcXZ2Wzh5vx6fmz4.1
Password: 835700
Speaker: Gautam Greddy
Title: TBD
Abstract: TBD
Time: 9am, November 13 2026, GMT+8
Location: Online
Zoom: https://nus-sg.zoom.us/j/88911975409?pwd=ixb7zXPuROkgclFcXZ2Wzh5vx6fmz4.1
Password: 835700
Speaker: Linhao Li
Title: TBD
Abstract: TBD
The seminar series in the fall semester of 2026 is over, and the next seminar will take place in the spring semester of 2027.