Noto-Workshop on Machine Learning and Physics 2025

Venue: Ikoinomura Noto Peninsula / Dates: December 9--11, 2025

Descriptions

In 2023, we organized two small-sized workshops in Japan:

1. Topology and Machine Learning: Workshop in Shonan 2023

2. A Focused Week of Physics for Machine Learning in Kyoto 2023

In 2024, we continued this tradition and enhanced the scope to a mid-sized workshop in Heidelberg:

3. Physics of Intelligence and Machine Learning 2024

These workshops were extremely successful with active discussions in a friendly atmosphere.  In particular, a collaboration was spontaneously formed during the workshops, and productive outputs resulted from new ideas brought by cross-disciplinary backgrounds.

What is the most effective way to maximize achievement performance?  We believe that one answer is to collect as strong researchers as possible to one place, and confine them with ample discussion time for a couple of days.  This format of meeting is called "Gassyuku" in Japanese (literal translation is "Camp"), which is a sort of good traditional event to tackle hard obstacles intensely with a small group of people.

This workshop or Gassyuku will have lecturers and the participants can continue discussions as long as the conference room is open.  We already emphasized this in the previous event, but again, we would like to rephrase the importance of physics in the development of machine-learning technologies.  The Nobel-Prize press-release states: They trained artificial neural networks using physics.  Indeed, physics is such a universal language, and we are hoping to find more and more crossings between physics and other fields.

Participation is by invitation only.
If you are interested, please contact the organizer (Kenji Fukushima).

Invited Speakers and Talk Titles

• Toshihiro Ota (太田敏博) (CyberAgent AI Lab)
  --- Attention Mechanism from Scratch

• Takashi Takahashi (高橋昂) (U. Tokyo)
  --- Variational Replica Analysis of Learning (学習の変分レプリカ解析)

• Sho Yokoi (横井祥) (NINJAL)
  --- Natural Language as a Geometric Object (言語の表現学習と意味の形)

• Akinori Tanaka (田中章詞) (RIKEN-AIP)
  --- Towards ELBO in Quantum Information/Machine Learning

• Yuuki Nakaguchi (中口悠輝) (NEC)
  --- Maximum Entropy Reinforcement Learning

• Lingxiao Wang (RIKEN/U. Tokyo))
  --- Learning Phase Transitions from Macroscopic to Microscopic Perspectives

• Jan Pawlowski (Heidelberg U.)

Participants

• Wenhua Cai (U. Tokyo)

• Kenji Fukushima (U. Tokyo)

• Yuji Hirono (Tsukuba U.)

• Renzo Kapust (Heidelberg U.)

• Syo Kamata (U. Tokyo)

• Jinyang Li (KEK/RIKEN)

• Jie Mei (UCAS)

• Josuke Minamiguchi (U. Tokyo)

• Tatsuhiro Misumi (Kindai U.)

• Ken Shiozaki (YITP)

• Tomoya Uji (U. Tokyo)

• Hironao Yamato (Osaka Univ.)

• Genki Yoshimura (Osaka Univ.)

• Dong-Lin Wang (USTC/U. Tokyo)