Statphys Seminar June 16 by Dr. Keiichi Tamai

6/16

t-okubo@phys.s.u-tokyo.ac.jp

Zoom

——
StatPhys Seminar @ UTokyo Hongo
sites.google.com/view/statphys-seminar

Room: 1206 (Room No. 206, Science 1st Bldg.) and Online(Zoom)

Speaker: , Keiichi Tamai (Institute for Physics of Intelligence, The University of Tokyo)

Title: Universal Scaling Laws of Absorbing Phase Transitions in Complex Systems

Abstract:
The notion of universality in critical phenomena, which is well-known for equilibrium phase transitions, can be extended to non-equilibrium ones [1,2]. While non-equilibrium critical phenomena were primarily of theoretical interest in the last century, recent progress suggests that they may be relevant for a deeper understanding of practically important complex systems. In this talk, I will focus on absorbing phase transitions (transitions to a state from which systems cannot escape) to see this point in more detail. After a brief recap on the scaling theory for non-equilibrium critical phenomena, I will demonstrate how universal scaling laws can be seen in various complex systems; in particular, open shear flows in the transitional regime [3,4] and classical artificial deep neural networks near the edge of chaos [5].

[1] H. Hinrichsen. Adv. Phys. 49, 815 (2000).
[2] M. Henkel, H. Hinrichsen, S. Lbeck. Non-equilibrium Phase Transitions. Vol. 1 (Springer, 2008).
[3] M. Sano & KT. Nat. Phys. 12, 249 (2016).
[4] K. Kohyama, M. Sano, KT & T. Tsukahara. Proceedings of TSFP-12 (2022).
[5] KT, T. Okubo, T. V. T. Duy, N. Natori & S. Todo. Under review.
———
##############################################

113-00337-3-1

19 950
e-mail: t-okubo@phys.s.u-tokyo.ac.jp
Tel: 03-5841-8890

————————————————-
Computational Material Physics Mailing List
home: www.issp.u-tokyo.ac.jp/public/cmp/
archive: cmp-ml.issp.u-tokyo.ac.jp
twitter: https://twitter.com/cmp_ml
————————————————-

————————————————-
Computational Material Physics Mailing List
home: www.issp.u-tokyo.ac.jp/public/cmp/
archive: cmp-ml.issp.u-tokyo.ac.jp
twitter: https://twitter.com/cmp_ml
————————————————-