AI 中文总结
本研究针对自发对称破缺机（SSBM），明确其类伊辛机的驱动条件限制，通过添加类物理退火效应的伪退火驱动方法，实现了K₂₀₀₀基准问题的最佳割值搜索。
AI 中文摘要
在我们之前的论文中，我们描述了一种新型物理实现型模拟器（自发对称破缺机，SSBM）的提出与验证，该模拟器用于组合优化问题，利用了由自发对称破缺产生伪自旋多体系统的模型的对偶现象。此外，我们在大规模基准问题K₂₀₀₀上针对不同初始涨落进行了数值模拟，报告称我们确定了一种条件，在此条件下，伪自旋模式在解搜索过程中逐渐收敛，最终形成单一模式，且该收敛解的割值达到已知最佳值的99.7%。在本文中，我们讨论了SSBM可表现出类连续变量型伊辛机行为的条件，并报告称，通过在这些条件中添加类似物理退火的效应，可使SSBM搜索K₂₀₀₀中的已知最佳割。
英文摘要
In our previous paper, we described the proposal and validation of a new physical implementation-type simulator (spontaneous symmetry breaking machine (SSBM)) for a combinatorial optimization problem that utilizes a phenomenon dual to the model in which a pseudo-spin many-body system is created by spontaneous symmetry breaking. Furthermore, we performed numerical simulations with different initial fluctuations on a large-scale benchmark problem (K_{2000}) and reported that we identified a condition under which pseudo-spin patterns gradually converge during the solution search process to ultimately form a single pattern and that the cut value of this converged solution reached 99.7% of the known best. In this paper, we discuss the conditions under which the SSBM can exhibit behavior analogous to that of a continuous-variable-type Ising machine, and report that by adding an effect similar to physical annealing to these conditions, the SSBM can be made to search for the best known cuts in K_{2000}.
Comments5 page, 3 figures and 1 Supplementary Materials file
