Sheaf neural networks on the SPD manifold enable strictly more expressive second-order geometric representations than Euclidean versions and achieve SOTA results on most MoleculeNet benchmarks.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Sheaf Neural Networks on SPD Manifolds: Second-Order Geometric Representation Learning
Sheaf neural networks on the SPD manifold enable strictly more expressive second-order geometric representations than Euclidean versions and achieve SOTA results on most MoleculeNet benchmarks.