Neuromorphic architecture with multiple coupled neurons using internal state neuron information
AssigneeInternational Business Machines Corporation
InventorsAngeliki Pantazi, Tomas Tuma, Stanislaw Wozniak
Life of this patent
Estimating the term…
Abstract
This invention relates to an apparatus, system, and method for computing with neuromorphic circuit architectures that have neurons with interconnected internal state information. The interconnected internal state information allows the neurons to enable or strengthen the input to other neurons. Furthermore, neuron internal state information provides insights on the characteristics of the input data that can be used to enhance the performance of the neuromorphic system. The neuromorphic system can be implemented with an artificial phase-change-based neurons.
Claims20
An apparatus for computation, comprising:
- a. a first electronic neuron comprising a first internal state;
- and
- b. a second electronic neuron, wherein the first internal state is connected to an internal state input of the second electronic neuron, thereby modifying the operation of the second electronic neuron.
An apparatus for computation, comprising:
- a. a first electronic neuron comprising a first internal state;
- b. a second electronic neuron, wherein the first internal state is connected to an internal state input of the second electronic neuron, thereby modifying the operation of the second electronic neuron;
- and
- c. a third electronic neuron, wherein the first internal state is connected to an internal state input of the third electronic neuron, thereby modifying the operation of the third electronic neuron.
A method for computation, the method comprising:
- a. receiving a first set of inputs at a first spike-timing dependent plasticity synapse;
- b. receiving a second set of inputs at a second spike-timing dependent plasticity synapse;
- c. combining the output of the first spike-timing dependent plasticity synapse and the second spike-timing dependent plasticity synapse in a first neuron to generate a cumulative postsynaptic potential signal;
- and
- d. receiving the cumulative postsynaptic potential signal at an input of a second neuron and modifying the operation of the second neuron based on the cumulative postsynaptic potential signal.
Dependencies are inferred from the claim text.