activity
20202026
most citedPath Integral Based Convolution and Pooling for Graph Neural Networks

33 citations · 84 across the 11 of their papers we have counts for

collaborators
Showing 2026Show all

1 paper · 1 filter

cs.LG2026

ProtoGuide: Prototype-Driven Guidance for Class-Conditional Graph Generation

Salvatore Romano, Marco Grassia, Pietro Liò +1

Discrete diffusion models are a prominent family for graph generation, but standard class-conditional mechanisms embed the class signal in the denoiser during training, tying the c…