We devise a hierarchy of spectral graph invariants, generalising the adjacency spectra and Laplacian spectra, which are commensurate in power with the hierarchy of combinatorial graph invariants generated by the Weisfeiler--Leman (WL) algorithm. More precisely, we provide a spectral characterisation of $k$-WL indistinguishability after $d$ iterations, for $k,d \in \mathbb{N}$. Most of the well-known spectral graph invariants such as adjacency or Laplacian spectra lie in the regime between 1-WL and 2-WL. We show that individualising one vertex plus running 1-WL is already more powerful than all such spectral invariants in terms of their ability to distinguish non-isomorphic graphs. Building on this result, we resolve an open problem of F\"urer (2010) about spectral invariants and strengthen a result due to Godsil (1981) about commute distances.
翻译:我们设计了光谱图变量的等级, 概括了相邻光谱和拉普拉西亚光谱, 这些光谱与Weisfeiler- Leman(WL)算法生成的组合图变量的等级相对应。 更准确地说, 我们为 $k$, d\ in\ mathb{N} $ 提供了美元迭代后无差异的光谱特性。 大多数著名的光谱图变量, 如相邻光谱或拉普拉西亚光谱, 都位于1- WL 和 2 WL 之间。 我们显示, 单化一个顶点加上运行的1 WL 已经比所有这种光谱差异在区分非色变图的能力方面更加强大。 在此基础上, 我们解决了F\ urer( 2010) 有关光谱变量的公开问题, 并强化了因上帝( 1981) 而产生的关于通向距离的结果 。