Logic-based Weisfeiler-Leman variants enable graph-to-table conversion for classification that matches GNN and graph transformer accuracy while running 5-20x faster without GPUs.
The first is, for some1 ≤ i ≤ n, to move one pebble onto a node inN(v) \ X ′ i and the other onto a node inXi, then take the role of the defender
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Graph Learning via Logic-Based Weisfeiler-Leman Variants and Tabularization
Logic-based Weisfeiler-Leman variants enable graph-to-table conversion for classification that matches GNN and graph transformer accuracy while running 5-20x faster without GPUs.