PEARL generates expressive, stable, and scalable graph positional encodings by passing random or basis node features through message-passing GNNs and pooling the outputs.
Pf-gnn: Differentiable particle filtering based approximation of universal graph representations
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Learning Efficient Positional Encodings with Graph Neural Networks
PEARL generates expressive, stable, and scalable graph positional encodings by passing random or basis node features through message-passing GNNs and pooling the outputs.