Research map / Graph learning / GNN architectures
Node Dependent Local Smoothing for Scalable Graph Learning
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 |
|---|---|---|---|
| use node-dependent local smoothing for features and predictions | NDLS with MLP · Cora | test accuracy: 84.4$\pm$0.5 -> 84.6$\pm$0.5 | Helped |
| use node-dependent local smoothing for features and predictions | NDLS with MLP · Citeseer | test accuracy: 73.3$\pm$0.6 -> 73.7$\pm$0.6 | Helped |
| use node-dependent local smoothing for features and predictions | NDLS with MLP · PubMed | test accuracy: 80.6$\pm$0.4 -> 81.4$\pm$0.4 | Helped |
| use node-dependent local smoothing for features and predictions | NDLS with MLP · Industry | test accuracy: 47.2$\pm$0.3 -> 47.7$\pm$0.5 | Helped |
| use node-dependent local smoothing for features and predictions | NDLS with MLP · ogbn-papers100M | test accuracy: 65.2$\pm$0.3 -> 65.6$\pm$0.3 | Helped |
| use node-dependent local smoothing for features and predictions | NDLS with MLP · Flickr | test accuracy: 51.1$\pm$0.1 -> 52.6$\pm$0.4 | Helped |
| use node-dependent local smoothing for features and predictions | NDLS with MLP · Reddit | test accuracy: 96.6$\pm$0.1 -> 96.8$\pm$0.1 | Helped |
| use NDLS instead of GAT | node classification · ogbn-arxiv | test accuracy: 73.56 -> 73.04 | Hurt |
| add NDLS-L label smoothing (ablation) | NDLS-F+MLP · Cora | test accuracy: 84.1$\pm$0.6 -> 84.6$\pm$0.5 | Helped |
| add NDLS-L label smoothing (ablation) | NDLS-F+MLP · PubMed | test accuracy: 81.1$\pm$0.6 -> 81.4$\pm$0.4 | Helped |
| add NDLS-F feature smoothing (ablation) | MLP+NDLS-L · Industry | test accuracy: 46.9$\pm$0.7 -> 47.7$\pm$0.5 | Helped |
| add NDLS-L label smoothing (ablation) | NDLS-F+MLP · Reddit | test accuracy: 96.6$\pm$0.1 -> 96.8$\pm$0.1 | Mixed |
Closest papers by meaning
- Posterior Label Smoothing for Node Classification (AAAI 2026 · similarity 0.90)
- On Which Nodes Does GCN Fail? Enhancing GCN From the Node Perspective (ICML 2024 · similarity 0.90)
- NLGT: Neighborhood-based and Label-enhanced Graph Transformer Framework for Node Classification (AAAI 2025 · similarity 0.87)
- Smoothness Really Matters: A Simple Yet Effective Approach for Unsupervised Graph Domain Adaptation (AAAI 2025 · similarity 0.87)
- Deep Graph Neural Networks via Posteriori-Sampling-based Node-Adaptative Residual Module (NeurIPS 2024 · similarity 0.87)
- Improving Graph Neural Networks by Learning Continuous Edge Directions (ICLR 2025 · similarity 0.87)
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.