Adding a target-derived prototype as an input improves GNN footprint emulator scores, but only under an oracle assignment that leaks the true label.
Evaluation of lagrangian particle dispersion models with measurements from controlled tracer releases
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
Prototype-enhanced prediction in graph neural networks for climate applications
Adding a target-derived prototype as an input improves GNN footprint emulator scores, but only under an oracle assignment that leaks the true label.