Research map / Graph learning / GNN architectures
Hierarchical Graph Transformer with Adaptive Node Sampling
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add adaptive node sampling to hierarchical Graph Transformer (ablation) | HGT · Chameleon | accuracy (%): 57.20 -> 65.42 | Helped |
| choose sampling heuristics adaptively rather than rely on one fixed heuristic | Graph Transformer on Newman networks · Newman networks with varying homophily | accuracy (%): 73.1%; 77.2% | Mixed |
| add coarse-grained global attention using coarsened super-nodes | ANS-GT · node classification datasets | accuracy (%) | Helped |
| vary the number of Transformer layers (ablation) | ANS-GT · Cora and other sensitivity-analysis datasets | accuracy | Mixed |
| vary the number of sampled super-nodes (ablation) | ANS-GT · Cora and other sensitivity-analysis datasets | accuracy | Mixed |
| vary the number of global nodes (ablation) | ANS-GT · Cora and other sensitivity-analysis datasets | accuracy | Mixed |
| increase the number of data augmentations and bagging samples (ablation) | ANS-GT · Cora and other sensitivity-analysis datasets | accuracy | Helped |
| vary the coarsening rate from c=0.01 toward larger values (ablation) | ANS-GT with VN coarsening · Cora | accuracy (%): 88.60 -> 87.26 | Hurt |
| replace VN graph coarsening with VE or JC (ablation) | ANS-GT · Cora and Actor | accuracy (%): Cora: 88.60, 87.95, 88.49; Actor: 39.72, 39.20, 39.15 | No clear effect |
Closest papers by meaning
- HubGT: Fast Graph Transformer with Decoupled Hierarchy Labeling (NeurIPS 2025 · similarity 0.90)
- Less is More: on the Over-Globalizing Problem in Graph Transformers (ICML 2024 · similarity 0.89)
- What Improves the Generalization of Graph Transformers? A Theoretical Dive into the Self-attention and Positional Encoding (ICML 2024 · similarity 0.89)
- Generalizable Insights for Graph Transformers in Theory and Practice (NeurIPS 2025 · similarity 0.89)
- Towards Principled Graph Transformers (NeurIPS 2024 · similarity 0.89)
- Unifying and Enhancing Graph Transformers via a Hierarchical Mask Framework (NeurIPS 2025 · similarity 0.89)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.