Research
On-device research index

arXiv research

A locally-built, LLM-digested index of recent arXiv papers in quant finance, geometry/topology, and statistical ML — keyword search served straight from SQLite on this machine.

169,341 papers · 148 categories

Trend · papers per month

71142213284 · Jun 202019922001200920182026
48 results for graph counting

Graph neural networks struggle with counting certain substructures in graphs.

problem Detecting and counting specific substructures in graphs.
method Study of graph neural networks' ability to count attributed graph substructures.
result Graph neural networks like MPNNs, 2-WL, and 2-IGNs have limitations in counting certain substructures.

New bootstraps improve speed and accuracy for graph count functionals.

problem Efficiently counting subgraphs in large graphs.
method Developed two types of multiplier bootstraps: a fast, approximate linear one and a quadratic one for denser graphs.
result Both bootstraps provide valid inference and higher-order accuracy under different graph sparsity conditions.

The paper develops formulas to count sizes of Markov equivalence classes of DAGs.

problem Measuring uncertainty and complexity in causal learning from DAGs.
method Introducing core graphs and deriving polynomial size formulas via symbolic computation.
result Efficient formulas for counting sizes of Markov equivalence classes of DAGs.

Introduces Niebrzydowski algebras for trivalent spatial graphs and handles.

problem Counting and distinguishing trivalent spatial graphs and handlebody-links.
method Defines Niebrzydowski algebras with ternary operation and partially defined multiplication, motivated by Reidemeister moves.
result Niebrzydowski algebras can distinguish some trivalent spatial graphs and handlebody-links.

This study compares GNNs and GA-MLPs, finding GA-MLPs can distinguish graphs but not count walks.

problem Comparing expressive power and graph isomorphism testing capabilities of GNNs and GA-MLPs.
method GA-MLPs augment node features with multi-hop operators and apply MLPs node-wise; GNNs are compared as a baseline.
result GA-MLPs can distinguish almost all non-isomorphic graphs but cannot count attributed walks, unlike GNNs.

Counting lattice points in moduli space of Klein surfaces.

problem Count lattice points in moduli space of Klein surfaces.
method Introduced metric Möbius graphs, counted lattice points weighted by non-orientability measure, deduced recursion for volumes.
result Proved refined version of Norbury's recursion and computed refined Euler characteristic.

New polynomials defined for quandle structures, enhancing graph invariants.

problem Enhancing the counting invariant for spatial graphs and handlebody-links.
method Introducing quandle polynomials and G-family polynomials for quandles, defining enhancements for invariants.
result New enhancements of the G-family counting invariant for trivalent spatial graphs and handlebody-links.

GraphMoE generates random graphs using neural networks and graphlets.

problem Learning generative models for random graphs.
method GraphMoE uses a neural network trained with graphlets and subgraph counts to match the distribution of random graphs.
result GraphMoE can generate graphs that mimic various real-world datasets and fool graph classifiers.

Graph Substructure Networks (GSN) improves GNN expressivity by counting subgraph isomorphisms.

problem Limited expressivity of GNNs in detecting and counting graph substructures.
method Topologically-aware message passing scheme based on substructure encoding.
result GSN is strictly more expressive than the Weisfeiler-Leman (WL) test and can disambiguate even hard graph isomorphism instances.

Efficient algorithm for graph matching in correlated stochastic block models.

problem Graph matching in correlated stochastic block models with balanced communities.
method Extends previous work on centered subgraph counts to handle estimation errors and edge correlation.
result First efficient algorithm for graph matching in the logarithmic average degree regime, matching all but a vanishing fraction of vertices with high probability.

A new test statistic counts tree co-occurrences to detect edge correlation between networks.

problem Detecting edge correlation between networks using latent vertex correspondence.
method The test statistic is based on counting co-occurrences of signed trees for a family of non-isomorphic trees.
result The test runs in n2+o(1)n^{2+o(1)} time and succeeds with high probability for large nn.

Study counts and equidistributes geodesic orbits on curved spaces.

problem Counting and equidistribution of strongly reversible closed geodesics in negatively curved spaces.
method Generalized techniques from Sarnak and Erlandsson-Souto, thermodynamic formalism, and graphs of groups with 2-torsion.
result Asymptotic counting and equidistribution of geodesic orbits towards the Bowen-Margulis measure.

Study shows central limit theorem for counting measures in non-smooth spaces.

problem Counting measures in non-smooth spaces with coarse negative curvature.
method Established central limit theorems for actions of groups on hyperbolic spaces without properness or smoothness assumptions.
result General framework allows for applications in geometrically finite manifolds and intersection numbers.

R-GPM enables efficient graph pattern mining through user-defined relations.

problem Efficient graph pattern mining through user-defined relations.
method Parallel computing framework with MCMC sampling algorithm and optimizations.
result Efficient estimators for graph pattern statistics with up to 3-orders-of-magnitude computational cost reduction.

Proves quasimodularity of generating functions for pillowcase covers.

problem Counting Feynman-like graphs associated with quadratic differentials.
method Analyzing decompositions of half-translation surfaces into horizontal cylinders.
result Alternative proof of quasimodularity results and practical method to compute area Siegel-Veech constants.

Study uses knot theory to model RNA foldings, emphasizing both entanglement and intrachain interactions.

problem Modeling RNA foldings considering both entanglement and intrachain interactions.
method Combines knot theory with embedded rigid vertex graphs to emphasize both entanglement and intrachain interactions of RNA foldings.
result Defines and computes a coloring counting invariant for stuck links, providing explicit computations for arc diagrams of RNA foldings.

We find the minimal number of links in an embedding of any complete kk-partite graph on 7 vertices (including K7K_7, which has at least 21 links). We give either exact values or upper and lower bounds for the minimal number of links for all complete kk-partite graphs on 8 vertices. We also look at larger complete bip…

2006-11-21abs ↗pdf ↗