Research map / Graph learning / GNN architectures
Graph Neural Networks can Recover the Hidden Features Solely from the Graph Structure
Ryoma Sato
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 |
|---|---|---|---|
| Replace explicit features with GNN-recovered hidden features (vs. synthetic degree/size features only) | Node classification via logistic regression · Cora, CiteSeer, PubMed, Coauthor CS/Physics, Computers, Photo | accuracy: e.g., Cora 0.122 -> 0.671; Coauthor CS 0.066 -> 0.492 | Improved |
| Use baseline GIN/GAT on synthetic features for hidden-feature recovery (vs. proposed threshold-based GNN) | Two-moon, Adult feature recovery · Transductive and inductive two-moon/Adult | d_G recovery error | Hurt |
| Use standard message-passing GNN architecture (GIN/GAT) instead of the constructed scale-aware architecture | Hidden feature recovery · Two-moon and Adult, transductive | qualitative d_G recovery | Hurt |
| Use tSNE on explicit synthetic features instead of GNN recovery | Feature recovery · Two-moon, Adult | qualitative recovery | Hurt |
Closest papers by meaning
- How Powerful are Graph Neural Networks? (ICLR 2019 · similarity 0.90)
- On Vanishing Gradients, Over-Smoothing, and Over-Squashing in GNNs: Bridging Recurrent and Graph Learning (NeurIPS 2025 · similarity 0.90)
- Graph Neural Networks Use Graphs When They Shouldn't (ICML 2024 · similarity 0.90)
- How Graph Neural Networks Learn: Lessons from Training Dynamics (ICML 2024 · similarity 0.89)
- On the Interplay between Graph Structure and Learning Algorithms in Graph Neural Networks (ICML 2025 · similarity 0.88)
- What do Graph Neural Networks learn? Insights from Tropical Geometry (NeurIPS 2024 · similarity 0.88)
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.
