Research map / Graph learning / GNN architectures
BernNet: Learning Arbitrary Graph Spectral Filters via Bernstein Approximation
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 |
|---|---|---|---|
| replace GPR-GNN polynomial filter with Bernstein-basis filter | propagation layer · Chameleon | accuracy (%): 67.28 -> 68.29 | Helped |
| replace GPR-GNN polynomial filter with Bernstein-basis filter | propagation layer · Actor | accuracy (%): 39.92 -> 41.79 | Helped |
| replace GPR-GNN polynomial filter with Bernstein-basis filter | propagation layer · Squirrel | accuracy (%): 50.15 -> 51.35 | Helped |
| replace GPR-GNN polynomial filter with Bernstein-basis filter | propagation layer · Computers | accuracy (%): 86.85 -> 87.64 | Helped |
| replace GPR-GNN polynomial filter with Bernstein-basis filter | propagation layer · Cora | accuracy (%): 88.57 -> 88.52 | No clear effect |
| replace GPR-GNN polynomial filter with Bernstein-basis filter | propagation layer · Photo | accuracy (%): 93.85 -> 93.63 | No clear effect |
| replace GPR-GNN polynomial filter with Bernstein-basis filter | propagation layer · Texas | accuracy (%): 92.95 -> 93.12 | No clear effect |
| use BernNet instead of ChebNet to learn comb filter | filter learner · Comb filter on images | sum of squared error: 4.0735 -> 0.9982 | Helped |
| use BernNet instead of GPR-GNN to learn band-pass filter | filter learner · Band-pass filter on images | sum of squared error: 3.5121 -> 0.0411 | Helped |
| use BernNet instead of GPR-GNN | model · per-epoch training time | time (ms): 9.94 -> 19.71 ms/epoch on Cora | Hurt |
Closest papers by meaning
- Unifying Homophily and Heterophily for Spectral Graph Neural Networks via Triple Filter Ensembles (NeurIPS 2024 · similarity 0.93)
- HoloNets: Spectral Convolutions do extend to Directed Graphs (ICLR 2024 · similarity 0.90)
- Position: Spectral GNNs Rely Less on Graph Fourier Basis than Conceived (ICML 2025 · similarity 0.90)
- SLOG: An Inductive Spectral Graph Neural Network Beyond Polynomial Filter (ICML 2024 · similarity 0.90)
- How Universal Polynomial Bases Enhance Spectral Graph Neural Networks: Heterophily, Over-smoothing, and Over-squashing (ICML 2024 · similarity 0.90)
- A General Graph Spectral Wavelet Convolution via Chebyshev Order Decomposition (ICML 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.