For every fixed k ≥ 2 the cyclic attractor detection problem is NP-complete precisely when the local Boolean function class contains majority-like self-dual rules or mixed conjunctive-disjunctive monotone families, and polynomial-time solvable in all other Post classes.
super hub
The Structure and Function of Complex Networks
12 Pith papers cite this work, alongside 13,819 external citations. Polarity classification is still indexing.
12
Pith papers citing it
13.8k
external citations · Crossref
hub tools
citation-role summary
method 1
citation-polarity summary
claims ledger
- method are present), as well as centrality measures (degree [22], betweenness [23], closeness [22] for each node). We averaged these across nodes to compare overall connectedness. We also calculated clustering coefficients (local and global) and counted simple motifs (triangles) [25, 26, 27]. To see how factors were grouped, we ran three community detection algorithms - Leiden [28], Girvan-Newman [29], and Infomap [30], on each graph. We analyzed whether communities contained nodes of the same 5P categ
authors
co-cited works
fields
cs.SI 2 cs.CC 1 cs.CL 1 cs.DS 1 cs.LG 1 econ.TH 1 physics.soc-ph 1 q-bio.NC 1 q-bio.PE 1 q-bio.QM 1years
2026 12verdicts
UNVERDICTED 12roles
method 1polarities
use method 1representative citing papers
Hybrid sketching saves up to 97% space on dense graphs and 15% on sparse ones by sketching dense cores and storing sparse parts exactly, with new BalloonSketch reducing sketch sizes up to 8x.
Network structures of applicant-vacancy links determine matching function forms, with dispersion in search intensities reducing match efficacy and potentially making higher average search counterproductive.
A new modularity function Qbg allows detection of hierarchical communities in bipartite networks at multiple scales by exploiting resolution limits.
Local 2- and 3-cycles enhance RNN computational capacity for Boolean functions, predicted by structural statistics, while adding interneurons boosts large networks.
Constructs hypergraphs from caCOH multivariate connectivity for EEG/MEG, recovering simulated coupling frequencies better than MSC graphs with reduction from 610 edges to 10 or 1 hyperedges.
A mechanistic model with horizontal gene transfer, new gene capture, genome emergence, and gene loss generates scale-free gene degrees and exponential genome degrees in bipartite networks, closely matching viral and pangenome observations when gene loss rate is set to zero.
Vehicle trips exhibit three phases explained by time-minimizing movement on hierarchical road networks.
IO-aware GPU kernels for SpMM convolutions, degree-aware reductions, and fused attention layers deliver median speedups of 1.6-2.6x (up to 10x) and memory reductions up to 76x over DGL/PyG baselines on realistic graphs.
Generalizes L-Modularity into a unified framework for community detection in complex link streams that include delayed, directed, weighted, and multipartite interactions.
A convolution process on a directed network provides a covariance model for SST that respects physical barriers and currents, used to identify thermal hot spots via Monte Carlo RCP projections.
LLMs generate 5P causal graphs from 46 psychotherapy intake transcripts that match human expert graphs in structure and meaning, with moderate clinical usefulness ratings.
citing papers explorer
-
On Efficient Scaling of GNNs via IO-Aware Layers Implementations
IO-aware GPU kernels for SpMM convolutions, degree-aware reductions, and fused attention layers deliver median speedups of 1.6-2.6x (up to 10x) and memory reductions up to 76x over DGL/PyG baselines on realistic graphs.