Introduces low-rank gluings in TLNs proving global fixed points are combinations of local fixed points, with complete characterization for rank-1 case and extension to gCTLNs.
On graphical domination for threshold-linear networks with recurrent excitation and global inhibition
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
fields
q-bio.NC 3years
2026 3verdicts
UNVERDICTED 3representative citing papers
Local 2- and 3-cycles enhance RNN computational capacity for Boolean functions, predicted by structural statistics, while adding interneurons boosts large networks.
Sequential chaotic oscillations arise in E-I threshold-linear networks under constant input, with transition order predictable from the graph when singleton fixed points are unstable and inhibition is strong.
citing papers explorer
-
Fixed point compositionality via low-rank gluing rules in inhibition-dominated threshold-linear networks
Introduces low-rank gluings in TLNs proving global fixed points are combinations of local fixed points, with complete characterization for rank-1 case and extension to gCTLNs.
-
Identifying structural design principles shaping the computational abilities of recurrent neural networks
Local 2- and 3-cycles enhance RNN computational capacity for Boolean functions, predicted by structural statistics, while adding interneurons boosts large networks.
-
Sequential chaotic oscillations in excitatory-inhibitory threshold-linear networks
Sequential chaotic oscillations arise in E-I threshold-linear networks under constant input, with transition order predictable from the graph when singleton fixed points are unstable and inhibition is strong.