GSWL test bounds the power of geometry-aware simplicial message passing, can be matched by such networks on finite families, and together with the Euler Characteristic Transform yields a complete geometric expressivity characterization.
Title resolution pending
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
fields
cs.LG 3years
2026 3verdicts
UNVERDICTED 3representative citing papers
GNNs and HOMP models saturate an extended manifold triangulation benchmark when given appropriate representations but show no generalization beyond combinatorial structure, indicating a gap in topology-aware learning.
Graph invariants serve as expressive, task-agnostic baselines that characterize structural heterogeneity and match trained models across 26 datasets, indicating that expressivity is not the primary driver of performance.
citing papers explorer
-
Geometry-Aware Simplicial Message Passing
GSWL test bounds the power of geometry-aware simplicial message passing, can be matched by such networks on finite families, and together with the Euler Characteristic Transform yields a complete geometric expressivity characterization.
-
No Triangulation Without Representation: Generalization in Topological Deep Learning
GNNs and HOMP models saturate an extended manifold triangulation benchmark when given appropriate representations but show no generalization beyond combinatorial structure, indicating a gap in topology-aware learning.
-
Invariant-Based Diagnostics for Graph Benchmarks
Graph invariants serve as expressive, task-agnostic baselines that characterize structural heterogeneity and match trained models across 26 datasets, indicating that expressivity is not the primary driver of performance.