Tokenphormer combines walk-tokens, SGPM-tokens, and hop-tokens in a graph transformer and reports improved node classification accuracy on several benchmark graphs.
Our experiments also verified this point: when the number of walks increased to a certain degree, the experimental results tended to be stable, and the model reached the fitting
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2024 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Tokenphormer: Structure-aware Multi-token Graph Transformer for Node Classification
Tokenphormer combines walk-tokens, SGPM-tokens, and hop-tokens in a graph transformer and reports improved node classification accuracy on several benchmark graphs.