People by Discipline

Name	Discipline	Category	Phone	Office
Mandelzweig Victor	Nuclear Hadronic and Few-Body Physics	Emeriti	02 6584 675	Kaplun 201
Unna Issachar	Nuclear Hadronic and Few-Body Physics	Emeriti	02 6584 806	Kaplun 104
Barnea Nir	Nuclear Hadronic and Few-Body Physics	Faculty	02 6585 138	Popick 114
Kirscher Johannes	Nuclear Hadronic and Few-Body Physics	Post-Docs		Ross 1C
Elyahu Moti	Nuclear Hadronic and Few-Body Physics	Graduate Students		Popick Bldg.
Krief Menahem	Nuclear Hadronic and Few-Body Physics	Graduate Students		Popick Bldg.
Gavrielov Noam	Nuclear Hadronic and Few-Body Physics	Graduate Students	02 6586 788	Kaplun 213
Paul Michael	Nuclear Hadronic and Few-Body Physics	Emeriti	02 6584 795	Kaplun 16
Panagiotis Georgoudis	Nuclear Hadronic and Few-Body Physics	Post-Docs	02 6584 354	Kaplun 202
Ohayon Ben	Nuclear Hadronic and Few-Body Physics	Graduate Students	02 6585 868	Kaplun Bldg.

Upcoming Events

Yesterday

Nonlinear Physics seminar: "Statistical mechanics of wide but finite artificial neural networks"
- Start time: 12:00pm
- End date: Wednesday, December 6th
- End time: 01:30pm
- Lecturer: Mr. Gadi Naveh (HUJI)Abstract: Over the past decade, Deep neural networks (DNNs) have brought significant advancements to various fields of science and technology. Key to their success is feature learning, enabling them to extract useful representations from raw data. However, DNNs remain challenging to interpret and lack a comprehensive analytical understanding. Recent studies revealed that DNNs in the infinite width limit are equivalent to Gaussian Processes (GPs), which are well-understoodanalytically but lack feature learning capabilities due to pre-determined features.In this thesis, we aim to bridge the gap between GPs and finite DNNs, using approaches inspired by theoretical physics. Employing statistical field theory, we treat the DNN output as a random object from a statistical ensemble. We performed extensive computer simulations of DNNs trained in both stylized and real-world settings. As for analytical methods, we first employ a perturbative approach to obtain the leading correction in 1/width to the GP limit. Subsequently, we develop a self-consistent mean-field type theory, providing a GP description of finite DNNs with a data-dependent kernel that embodies feature learning properties.
- View this event in Google Calendar
Position at Kyoto
- Start time: 02:00pm
- End date: Wednesday, December 6th
- End time: 03:00pm
- View this event in Google Calendar

Search form

People by Discipline

Pages