セミナー

Topological Image Analysis

2025年12月25日(木) 12:00 - 13:00

鍛冶 静雄 (九州大学 マス・フォア・インダストリ研究所 教授)

Topological Data Analysis (TDA) applies algebraic topology to the study of data such as point clouds. When applied to image and volumetric data, TDA provides a way to capture the topological features that characterise shapes and spatial structures. In this talk, I will outline the strengths and limitations of TDA for image analysis, and compare its capabilities with those of deep neural networks. I will also present hands-on examples using our open-source software Cubical Ripser. Finally, I will highlight a new direction in the use of TDA for image processing.

会場: SUURI-COOL (Kyushu) C501（九州大学 伊都キャンパス West 1号館 D棟） (メイン会場) / via Zoom

イベント公式言語: 英語

Cosmological correlators beyond the de-Sitter invariance

2025年12月24日(水) 15:00 - 17:00

Zhu Yuhang (Postdoctoral Researcher, Particle Theory and Cosmology Group, Center for Theoretical Physics of the Universe (CTPU), Institute for Basic Science (IBS), Republic of Korea)

Cosmological correlators serve as powerful probes of the physics that governed the Universe in its earliest moments. Yet analytic calculations of correlators involving massive spinning fields are highly challenging. Recent progress in the cosmological bootstrap program has greatly deepened our understanding of these correlators. In this talk, we will show how to extend the bootstrap program beyond exact de Sitter invariance by studying two types of symmetry breaking: explicit scale-invariance breaking and boost breaking. We will present the boundary differential equations that characterise correlators in these settings and highlight the rich phenomenology that emerges. Finally, we will also show recent developments in approximation methods, based on the combination of exact WKB and saddle-point method, which provide a precise, efficient, and physically transparent way to capture and classify the non-analytic features of correlators.

会場: via Zoom / セミナー室 (359号室)

イベント公式言語: 英語

Neural Network Quantum States for Quarkonium in Medium: Real-Time Open Quantum-System Dynamics

2025年12月19日(金) 16:00 - 17:00

Tom Magorsch (Ph.D. Student, Department of Physics, Technical University of Munich, Germany)

Many phenomena in high energy physics can not be described by Euclidean-time Monte Carlo estimates alone, but require genuine real-time evolution and a treatment of non-equilibrium effects. However, such simulations are computationally challenging. One such example is the evolution of heavy quarkonium in the quark gluon plasma produced in heavy-ion collisions. In this talk, I will introduce the open quantum system treatment of in-medium quarkonium. I will then give an overview on neural network quantum states as a variational approach to the real-time simulation of open quantum systems. As a controlled benchmark system, I will study the application to the Caldeira-Leggett model and conclude with an outlook on future applications of neural network based simulation of quarkonia in medium.

会場: via Zoom

イベント公式言語: 英語

Two-component dark matter scenario in a pseudo-Nambu-Goldstone dark matter model - Tomohiro Abe

2025年12月18日(木) 14:00 - 15:30

阿部 智広 (名古屋大学 素粒子宇宙起源研究所 (KMI) 特任助教)

The WIMP dark matter (or thermal dark matter) is one of the leading candidates for dark matter and is widely studied. On the other hand, recent progress in direct detection experiments places severe constraints on WIMP dark matter models. Pseudo-Nambu-Goldstone boson dark matter models (pNG DM models) can explain the measured value of the dark matter energy density via the freeze-out mechanism and also naturally suppress the dark matter-nucleon scattering. The amplitudes for the processes are proportional to the momentum transfer squared from the dark sector to the visible sector; It is suppressed by t ~ 0 for the dark matter-nucleon scattering process, while it is proportional to s ~ 4 m_DM^2 for annihilation processes. The simplest realization faces the domain wall problem and needs to be extended. Several models have been proposed, but each model brings other issues, a large hierarchy in new energy scales, parameter tunings, and a Landau pole of a new gauge coupling. In this talk, I would like to propose a new pNG DM model based on SU(2)_gauge times SU(2)_global symmetry that overcomes those issues. The model predicts a two-component DM scenario, where one is pNG, and the other is an ordinary WIMP. I will show that the effective spin-independent cross section in the direct detection experiments is smaller than the current upper bound and larger than the prospect of the Darwin project in a wide range of parameter space in the two-component DM scenario.

会場: セミナー室 (359号室) 3階 359号室とZoomのハイブリッド開催

イベント公式言語: 英語

Origin and evolutionary history of an urban underground mosquito

2025年12月18日(木) 13:00 - 14:00

羽場 優紀 (Postdoc, Zuckerman Institute, Columbia University, USA)

Urbanization is rapidly reshaping landscapes around the world, which poses questions about whether and how quickly animals and plants can adapt. Culex pipiens form molestus, more commonly known as the "London Underground mosquito," has been held up as a benchmark for the potential speed and complexity of urban adaptation. This intraspecific lineage within Cx. pipiens, a major West Nile virus vector, is purported to have evolved human biting and a suite of other human-adaptive behaviors in the subways and cellars of northern Europe within the past 200 years. Form molestus features prominently in textbooks as well as scholarly reviews of urban adaptation. Yet, the hypothesis of in situ urban evolution has never been rigorously tested. I will talk our recent efforts to understand the contentious origin and evolutionary history of the urban, human-biting mosquito. Our synthesis and meta-analysis of rich yet confusing literature show that its London Underground origin is unlikely (Haba and McBride 2022 Current Biology). Whole genome resequencing and population genomics of 800+ mosquitoes across ~50 countries again debunk the in situ evolution hypothesis and instead support that molestus first adapted to human environments >1000 years ago in the Mediterranean or Middle East, most likely in ancient Egypt or another early agricultural society (Haba et al. 2025 Science). I will outline implications of our results in urban evolutionary biology as well as in public health. Speaker Bio Yuki Haba, Ph.D., is an evolutionary biologist passionate about understanding how and why diverse behaviors evolve in nature. He is currently a Leon Levy Scholar in Neuroscience at Columbia University's Zuckerman Mind Brain Behavior Institute. He aims to take multi-desciplinary approaches, combining genomics, neuroscience, and field-based behavioral ecology to comprehensively understand the evolution of behavior. Yuki completed his PhD at Princeton, MA at Columbia, and undergraduate degree at the University of Tokyo. Personal webpage: https://yukihaba.github.io/

会場: セミナー室 (359号室)

イベント公式言語: 英語

A bi-fidelity Asymptotic-Preserving Neural Network approach for multiscale kinetic problems

2025年12月17日(水) 11:00 - 12:00

Liu Liu (Assistant Professor, Department of Mathematics, The Institute of Mathematical Sciences, The Chinese University of Hong Kong, Hong Kong)

In this talk, we will introduce a bi-fidelity Asymptotic-Preserving Neural Network (BI-APNNs) framework, designed to efficiently solve forward and inverse problems for the linear Boltzmann equation. Our approach builds upon the previously studied Asymptotic-Preserving Neural Network (APNNs), which employs a micro-macro decomposition to handle the model’s multiscale nature. We specifically address a bottleneck in the original APNNs: the slow convergence of the macroscopic density in the near fluid-dynamic regime. This strategy significantly accelerates the training convergence as well as improves the accuracy of the forward problem solution, particularly in the fluid-dynamic limit. We show several numerical experiments on both linear Boltzmann and the Boltzmann-Poisson system that this new BI-APNN method produces more accurate and robust results for forward and inverse problems compared to the standard APNNs. This is a joint work with Zhenyi Zhu and Xueyu Zhu.

会場: セミナー室 (359号室) (メイン会場) / via Zoom

イベント公式言語: 英語

Phase transition in parametrized quantum circuits

2025年12月16日(火) 10:00 - 12:00

王 啸洋 (理化学研究所 数理創造研究センター (iTHEMS) 数理展開部門 量子数理科学チーム 特別研究員)

会場: セミナー室 (359号室) (メイン会場) / via Zoom

イベント公式言語: 英語

Is the Fornax galaxy cluster hiding its radio sources?

2025年12月15日(月) 14:00 - 15:15

Alvina Yee Lian On (PD, Physics Division, National Center for Theoretical Sciences, Taiwan)

Recently, the ASKAP POSSUM and MeerKAT surveys revealed an apparent lack of radio source counts in the Fornax galaxy cluster field. The sources in this patch of sky also appeared to be less polarised. These observations are peculiar and could be important signatures of depolarisation on megaparsec scales. In this talk, I will present some of our recent results, where we quantified the effects on polarisation of radio point sources behind a merging galaxy cluster simulation. The merger gives rise to a large-scale intracluster shock, which compresses the gas and aligns the magnetic field along the shock front. The in-falling subcluster also triggers gas sloshing near the cluster center, resulting in the formation of cold fronts. Our ray-tracing calculations revealed that, generally, bright sources do not experience any significant changes in polarisation, whereas faint sources either get severely depolarised or enhanced as their radiation propagates through the intracluster medium. I will highlight the physical conditions under which polarisation may change significantly along the line-of-sight and discuss how these may impact our interpretations of radio observations, particularly with the SKA.

会場: セミナー室 (359号室) 3階 359号室とZoomのハイブリッド開催

イベント公式言語: 英語

RIKEN Quantum hands-on seminar on IBM and QEDMA software for quantum computing beginners

2025年12月15日(月) 9:30 - 12:20

This seminar will be conducted in a hybrid format, both in-person and via Zoom. Since it includes hands-on sessions, we kindly ask you to consider attending in person whenever possible to ensure more effective learning. The overview of the seminar is as follows: Program: 9:30 - 10:50 Yuri Kobayashi (IBM Quantum) "Introduction to IBM Quantum and Qiskit" 10:50 - 11:00 Break 11:00 - 12:20 Ori Alberton (QEDMA) "Introduction to QESEM error mitigation software: Theory,  use-case demonstrations and usage tutorial" Abstract of Yuri Kobayashi's tutorial: This talk introduces HPC researchers to the IBM Quantum platform and the Qiskit SDK, providing a practical orientation to quantum programming without assuming prior quantum-computing knowledge. Attendees will learn how to construct and execute quantum circuits using Qiskit, explore available simulators, and real quantum backends and submit jobs to IBM’s cloud-based quantum processors. The session will also showcase how to map your problem to quantum circuits through specific use-case applications. Abstract of Ori Alberton's tutorial: This talk introduces QESEM, Qedma’s characterization-based software tool for reliable, high-accuracy quasi-probabilistic error mitigation. We will begin by explaining why error mitigation methods are expected to be the first to unlock quantum advantage, and why they will continue to play a central role even as error correction becomes feasible. We will highlight some of the key innovations underlying QESEM’s operation and present recent results on IBM Heron r2 devices demonstrating its capabilities in the largest utility-scale unbiased error mitigation experiment to date. In addition different use-cases and demonstrations ran by QESEM users will be presented. Finally, we will provide guidance on how to start using QESEM to obtain error-mitigated results in experiments on IBM quantum systems.

会場: セミナー室 (359号室) (メイン会場) / via Zoom

イベント公式言語: 英語

Fracture squares and separable algebras

2025年12月12日(金) 16:00 - 17:30

Luca Pol (Postdoctoral Researcher, Max Planck Institute for Mathematics in Bonn, Germany)

In this talk I will present a way to reconstruct a category from its subcategories of complete and local objects while retaining the symmetric monoidal structure. As an application of this machinery I will discuss how to calculate separable algebras in equivariant homotopy theory.

会場: via Zoom / 研究本館 3階 セミナー室(345-347)

イベント公式言語: 英語

Widespread conservation of genetic effect sizes between human groups across traits

2025年12月12日(金) 13:30 - 15:00

Simon Robert Myers (Professor, University of Oxford, UK)

Understanding genetic differences between populations is essential for avoiding confounding in genome-wide association studies and improving polygenic score (PGS) portability. We developed a statistical pipeline to infer fine-scale Ancestry Components and applied it to UK Biobank data. Ancestry Components identify population structure not captured by widely used principal components, improving stratification correction for geographically correlated traits. To estimate the similarity of genetic effect sizes between groups, we developed ANCHOR, which estimates changes in the predictive power of an existing PGS in distinct local ancestry segments. ANCHOR infers highly similar (estimated correlation 0.98 ± 0.07) effect sizes between UK Biobank participants of African and European ancestry for 47 of 53 quantitative phenotypes, suggesting that gene–environment and gene–gene interactions do not play major roles in poor cross-ancestry PGS transferability for these traits in the United Kingdom, and providing optimism that shared causal mutations operate similarly in different populations.

会場: via Zoom / セミナー室 (359号室)

イベント公式言語: 英語

From perturbations of operators to noncommutative condensers

2025年12月11日(木) 16:00 - 17:00

Dan Voiculescu (Professor, Department of Mathematics, University of California, Berkeley, USA)

A numerical invariant, the quasicentral modulus underlies the multivariable generalizations of the classical Weyl-von Neumann-Kuroda and Kato-Rosenblum theorems. There are also connections to the Kolmogorov-Sinai dynamical entropy. I will also point out some of the open problems. Recently a noncommutative analogy with condenser capacity in nonlinear potential theory is emerging, that provides a new perspective on the subject.

会場: via Zoom / 研究本館 3階 359号室

イベント公式言語: 英語