A structured multi-layer recurrent SNN with fixed sparse connectivity and three-factor local plasticity rules supports deep computation and competitive supervised learning without backpropagation.
On the computational power of winner-take-all,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.NE 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Scalable Learning in Structured Recurrent Spiking Neural Networks without Backpropagation
A structured multi-layer recurrent SNN with fixed sparse connectivity and three-factor local plasticity rules supports deep computation and competitive supervised learning without backpropagation.