AI RESEARCH

Weisfeiler-Leman Is Incomplete on Simple Spectrum Graphs, so Canonicalize Them

arXiv CS.LG

ArXi:2605.23446v1 Announce Type: new Graphs with a simple spectrum admit cubic-time isomorphism testing, yet we prove that for every natural number $k$, the $k$-Weisfeiler-Leman ($k$-WL) test cannot distinguish all non-isomorphic graphs with a simple spectrum. As the WL hierarchy upper-bounds the distinguishing power of widely-used Graph Neural Networks (GNNs), this incompleteness applies to all such GNNs, ruling out completeness for every $k$-WL-aligned GNN family. To close this gap, we