Zobrazeno 1 - 10
of 31
pro vyhledávání: '"Zohar Ringel"'
Autor:
Doruk Efe Gökmen, Sounak Biswas, Sebastian D. Huber, Zohar Ringel, Felix Flicker, Maciej Koch-Janusz
Publikováno v:
Nature Communications, Vol 15, Iss 1, Pp 1-8 (2024)
Abstract The physics of complex systems stands to greatly benefit from the qualitative changes in data availability and advances in data-driven computational methods. Many of these systems can be represented by interacting degrees of freedom on inhom
Externí odkaz:
https://doaj.org/article/0827e7b3a4ee4d5e8f9dd91f0acc0136
Publikováno v:
Nature Communications, Vol 14, Iss 1, Pp 1-12 (2023)
In the quest to understand how deep neural networks work, identification of slow and fast variables is a desirable step. Inspired by tools from theoretical physics, the authors propose a simplified description of finite deep neural networks based on
Externí odkaz:
https://doaj.org/article/a93f28685df14ab280c445e62e02b891
Publikováno v:
Machine Learning: Science and Technology, Vol 5, Iss 3, p 035048 (2024)
Physically informed neural networks (PINNs) are a promising emerging method for solving differential equations. As in many other deep learning approaches, the choice of PINN design and training protocol requires careful craftsmanship. Here, we sugges
Externí odkaz:
https://doaj.org/article/73708e484d9147e3895b04786bf70366
Publikováno v:
Physical Review Research, Vol 3, Iss 2, p 023034 (2021)
In the past decade, deep neural networks (DNNs) came to the fore as the leading machine-learning algorithms for a variety of tasks. Their rise was founded on market needs and engineering craftsmanship, the latter based more on trial and error than on
Externí odkaz:
https://doaj.org/article/c0712b2145504c6aa3d16b45c32247a8
Publikováno v:
Physical Review Research, Vol 2, Iss 4, p 043032 (2020)
The infamous sign problem leads to an exponential complexity in Monte Carlo simulations of generic many-body quantum systems. Nevertheless, many phases of matter are known to admit a sign-problem-free representative, allowing efficient simulations on
Externí odkaz:
https://doaj.org/article/33d13ded4ffe4a93abd4012dd304b534
Publikováno v:
Physical Review Research, Vol 2, Iss 3, p 033515 (2020)
The sign problem is a widespread numerical hurdle preventing us from simulating the equilibrium behavior of various problems at the forefront of physics. Focusing on an important subclass of such problems, bosonic (2+1)-dimensional topological quantu
Externí odkaz:
https://doaj.org/article/dbcdb913f6d145dd8c69bd3a1c5428ea
Autor:
Patrick M. Lenggenhager, Doruk Efe Gökmen, Zohar Ringel, Sebastian D. Huber, Maciej Koch-Janusz
Publikováno v:
Physical Review X, Vol 10, Iss 1, p 011037 (2020)
Recently, a novel real-space renormalization group (RG) algorithm was introduced. By maximizing an information-theoretic quantity, the real-space mutual information, the algorithm identifies the relevant low-energy degrees of freedom. Motivated by th
Externí odkaz:
https://doaj.org/article/5b8edeefd9f44a91b1a7c64467a40de6
Publikováno v:
Physical Review Letters, 127 (24)
Physical Review Letters 127, 240603 (2021)
Physical Review Letters 127, 240603 (2021)
Identifying the relevant coarse-grained degrees of freedom in a complex physical system is a key stage in developing powerful effective theories in and out of equilibrium. The celebrated renormalization group provides a framework for this task, but i
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::6582e71226045da9348d6a22cbbe5749
https://hdl.handle.net/20.500.11850/519628
https://hdl.handle.net/20.500.11850/519628
Publikováno v:
Physical Review E. 104
A recent line of works studied wide deep neural networks (DNNs) by approximating them as Gaussian Processes (GPs). A DNN trained with gradient flow was shown to map to a GP governed by the Neural Tangent Kernel (NTK), whereas earlier works showed tha
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::2afc09cef1155278999f0106737ee048
https://doi.org/10.1103/physreve.104.064301
https://doi.org/10.1103/physreve.104.064301
Publikováno v:
Physical Review E, 104 (6)
Real-space mutual information (RSMI) was shown to be an important quantity, formally and from a numerical standpoint, in finding coarse-grained descriptions of physical systems. It very generally quantifies spatial correlations, and can give rise to
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::f206073873972d2fa82361f3d5d2595a
https://hdl.handle.net/20.500.11850/519629
https://hdl.handle.net/20.500.11850/519629