More is different.
God is in the details.
NEWS
- Title
- Neural network approach to scaling analysis of critical phenomena
- Reference
- Physical Review E 107, 044128 (2023)
- DOI
- 10.1103/PhysRevE.107.044128
- Author
- 10.1103/PhysRevE.107.044128
- Ryosuke Yoneda and Kenji Harada
- Abstract
- Determining the universality class of a system exhibiting critical phenomena is one of the central problems in physics. There are several methods to determine this universality class from data. As methods to collapse plots onto scaling functions, polynomial regression, which is less accurate, and Gaussian process regression, which provides high accuracy and flexibility but is computationally expensive, have been proposed. In this paper, we propose a regression method using a neural network. The computational complexity is linear only in the number of data points. We demonstrate the proposed method for the finite-size scaling analysis of critical phenomena in the two-dimensional Ising model and bond percolation problem to confirm the performance. This method efficiently obtains the critical values with accuracy in both cases.
- Comments
- 10 pages, 10 figures
- Preprint
- arXiv.2209.01777
日程: 2023年3月22日から3月25日
- 講演(22aL2-7) "テンソル化深層学習における圧縮性能の解析"
- 講演(23aL1-6) "ネットワーク構造最適化を含んだツリーテンソルネットワーク法の開発II" (共同研究者:引原俊哉, 上田宏, 奥西巧一, 西野友年)
- 講演(25pL1-9) "2次元横磁場イジング模型におけるグラウバーダイナミクスと動的量子臨界現象" (共同研究者:堀田知佐)
- Title
- Quantum critical dynamics in the two-dimensional transverse Ising model
- Reference
- Physical Review Research 5, 013186 (2023)
- DOI
- 10.1103/PhysRevResearch.5.013186
- Author
- 10.1103/PhysRevResearch.5.013186
- Chisa Hotta, Tempei Yoshida, and Kenji Harada
- Abstract
- In the vicinity of the quantum critical point (QCP), thermodynamic properties diverge toward zero temperature governed by universal exponents. Although this fact is well known, how it is reflected in quantum dynamics has not been addressed. The QCP of the transverse Ising model on a triangular lattice is an ideal platform to test the issue, since it has an experimental realization, the dielectrics being realized in an organic dimer Mott insulator, κ−ET2X, where a quantum electric dipole represents the Ising degrees of freedom. We track the Glauber-type dynamics of the model by constructing a kinetic protocol based on the quantum Monte Carlo method. The dynamical susceptibility takes the form of the Debye function and shows a significant peak narrowing in approaching a QCP due to the divergence of the relaxation timescale. It explains the anomaly of dielectric constants observed in the organic materials, indicating that the material is very near the ferroelectric QCP. We disclose how the dynamical and other critical exponents develop near QCP beyond the simple field theory.
- Comments
- 12 pages, 8 figures
- Preprint
- arXiv:2209.11599
- Title
- Quantum Circuit Simulation by SGEMM Emulation on Tensor Cores and Automatic Precision Selection
- Reference
- ISC 2023
- Author
- Hiryuki Ootomo, Hidetaka Manabe, Kenji Harada, and Rio Yokota
- Abstract
- Quantum circuit simulation provides the foundation for the development of quantum algorithms and the verification of quantum supremacy. Among the various methods for quantum circuit simulation, tensor network contraction has been increasing in popularity due to its ability to simulate a larger number of qubits. During tensor contraction, the input tensors are reshaped to matrices and computed by a GEMM operation, where these GEMM operations could reach up to 90\% of the total calculation time. GEMM throughput can be improved by utilizing mixed-precision hardware such as Tensor Cores, but straightforward implementation results in insufficient fidelity for deep and large quantum circuits. Prior work has demonstrated that compensated summation with special care of the rounding mode can fully recover the FP32 precision of SGEMM even when using TF32 or FP16 Tensor Cores. The exponent range is a critical issue when applying such techniques to quantum circuit simulation. While TF32 supports almost the same exponent range as FP32, FP16 supports a much smaller exponent range. In this work, we use the exponent range statistics of input tensor elements to select which Tensor Cores we use for the GEMM. We evaluate our method on Random Circuit Sampling (RCS), including Sycamore's quantum circuit, and show that the throughput is 1.86 times higher at maximum while maintaining accuracy.
- Preprint
- arXiv:2303.08989 [quant-ph]
- Title
- Automatic structural optimization of tree tensor networks
- Reference
- Physical Review Research 5, 013031 (2023)
- DOI
- 10.1103/PhysRevResearch.5.013031
- Author
- 10.1103/PhysRevResearch.5.013031
- Toshiya Hikihara, Hiroshi Ueda, Kouichi Okunishi, Kenji Harada, and Tomotoshi Nishino
- Abstract
- Tree tensor network (TTN) provides an essential theoretical framework for the practical simulation of quantum many-body systems, where the network structure defined by the connectivity of the isometry tensors plays a crucial role in improving its approximation accuracy. In this paper, we propose a TTN algorithm that enables us to automatically optimize the network structure by local reconnections of isometries to suppress the bipartite entanglement entropy on their legs. The algorithm can be seamlessly implemented to such a conventional TTN approach as density-matrix renormalization group. We apply the algorithm to the inhomogeneous antiferromagnetic Heisenberg spin chain having a hierarchical spatial distribution of the interactions. We then demonstrate that the entanglement structure embedded in the ground-state of the system can be efficiently visualized as a perfect binary tree in the optimized TTN. Possible improvements and applications of the algorithm are also discussed.
- Comments
- 11 pages, 10 figures, 2 tables
- Preprint
- arXiv:2209.03196
生理研研究会 「第4回 力学系の視点からの脳・神経回路の理解」にて、 "テンソルネットワークによるニューラルネットワークモデルの圧縮"について 講演を行いました.
TOPICS
オンラインで学ぶモンテカルロ法
モンテカルロ法(マルコフ過程を用いた手法も含む)の基本的な事柄についての解説。
解説ページへACTIVITY
学術変革領域(A)極限宇宙の物理法則を創る-量子情報で拓く時空と物質の新しいパラダイム
ABOUT

原田健自
(
Kenji Harada
)
京都大学大学院情報学研究科
助教
harada.kenji.8e@kyoto-u.ac.jp
京都市左京区吉田本町 京都大学吉田キャンパス 総合研究8号館203号室
Map
(No.59)
統計物理学と情報論的視点を融合した最先端の計算手法とスーパーコンピュータのパワーを組み合わせ、相互作用する多体系と情報科学における未解決問題に先端的に取り組んでいます。