1 paper
Xudong Wang, Chris Ding, Tongxin Li +1
We introduce GLASS, a framework for graph-level anomaly detection (GLAD) that achieves robust cross-domain transferability through graph-language alignment on the unit hypersphere.…