Research map / Graph learning / GNN architectures
Equivariant Polynomials for Graph Neural Networks
Omri Puny, Derek Lim, Bobak Kiani, Haggai Maron, Yaron Lipman
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 |
|---|---|---|---|
| Add degree-6 polynomial features to PPGN++ (ablation) | PPGN++ · ZINC 12K | Test MAE: .076 \pm .003 -> .071 \pm .001 | Improved |
| Add degree-5 polynomial features to PPGN++ (ablation) | PPGN++ · ZINC 12K | Test MAE: .076 \pm .003 -> .072 \pm .005 | Improved |
| Add polynomial features at degrees 4, 5, or 6 (ablation) | GatedGCN · ZINC 12K | Test MAE: .265 \pm .015 -> .150 \pm .005 (degree 4); .138 \pm .003 (degree 5); .106 \pm .003 (degree 6) | Improved |
| Add degree-6 polynomial features instead of degree-5 features (ablation) | PPGN++ · ZINC-full | Test MAE: .020 \pm .001 -> .020 \pm .001 | No clear effect |
| Add degree-6 polynomial features instead of degree-5 features (ablation) | PPGN++ · Alchemy | Test MAE: .110 \pm .001 -> .109 \pm .001 | Improved |
| Add polynomial features of increasing degree (ablation) | PPGN++ and GatedGCN · Strongly regular graph dataset (SR) | failure rate for distinguishing non-isomorphic graphs | Improved |
Closest papers by meaning
- Are High-Degree Representations Really Unnecessary in Equivariant Graph Neural Networks? (NeurIPS 2024 · similarity 0.92)
- Beyond Weisfeiler-Lehman: A Quantitative Framework for GNN Expressiveness (ICLR 2024 · similarity 0.90)
- Homomorphism Expressivity of Spectral Invariant Graph Neural Networks (ICLR 2025 · similarity 0.90)
- Schur Nets: exploiting local structure for equivariance in higher order graph neural networks (NeurIPS 2024 · similarity 0.90)
- On the Expressive Power of Spectral Invariant Graph Neural Networks (ICML 2024 · similarity 0.90)
- Spectral Graph Neural Networks are Incomplete on Graphs with a Simple Spectrum (NeurIPS 2025 · similarity 0.90)
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.
