Sign and Basis Invariant Networks for Spectral Graph Representation Learning
Reviewed by Pith T0 review T1 audit T2 compute T3 formal T4 kernel pith:2OPBJGZUrecord.jsonopen to challenge →
read the original abstract
We introduce SignNet and BasisNet -- new neural architectures that are invariant to two key symmetries displayed by eigenvectors: (i) sign flips, since if $v$ is an eigenvector then so is $-v$; and (ii) more general basis symmetries, which occur in higher dimensional eigenspaces with infinitely many choices of basis eigenvectors. We prove that under certain conditions our networks are universal, i.e., they can approximate any continuous function of eigenvectors with the desired invariances. When used with Laplacian eigenvectors, our networks are provably more expressive than existing spectral methods on graphs; for instance, they subsume all spectral graph convolutions, certain spectral graph invariants, and previously proposed graph positional encodings as special cases. Experiments show that our networks significantly outperform existing baselines on molecular graph regression, learning expressive graph representations, and learning neural fields on triangle meshes. Our code is available at https://github.com/cptq/SignNet-BasisNet .
This paper has not been read by Pith yet.
Forward citations
Cited by 6 Pith papers
-
When Do Graph Foundation Models Transfer? A Data-Centric Theory
Using graphon limits for dense graphs, the authors decompose cross-domain output shifts for Lipschitz backbones into graph-specific finite-sample terms and a relabeling-invariant domain discrepancy, with stability res...
-
Canopy: A Heterograph Foundation Model for Metabolic Engineering
Frozen embeddings from a pretrained heterogeneous graph transformer over a 6.9M-node metabolic-engineering knowledge graph predict fermentation titers at R²=0.41, outperforming tabular baselines (R²=0.24).
-
MidSurfNet: Learnable Face Pairing and Interference Implicit Fields for Generalized Mid-surface Abstraction
MidSurfNet combines a neural face pairing module with an interference implicit field to generate generalized mid-surfaces, reporting 87.32% pairing accuracy and improved handling of multi-wall and self-matching cases ...
-
Teaching LLMs to See Graphs: Unifying Text and Structural Reasoning
GTLM injects graph-aware attention biases into LLMs using only 0.015% extra parameters, enabling native graph processing that matches 7B models with a 1B model on text-attributed graph benchmarks.
-
Adaptive Canonicalization with Application to Invariant Anisotropic Geometric Networks
Adaptive canonicalization selects input canonical forms by maximizing network predictive confidence to yield continuous symmetry-preserving models with universal approximation for equivariant geometric networks.
-
Feature Augmentation of GNNs for ILPs: Local Uniqueness Suffices
Local d-hop uniqueness in GNN node features matches global UID expressiveness for ILP solving while providing stronger generalization.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.