US 2026/0277546 A1 · SYSTEM, METHOD, AND APPARATUS FOR RECURRENT NEURAL NETWORKS

Abstract

A method for computation with recurrent neural networks includes receiving an input drive and a recurrent drive, producing at least one modulatory response; computing at least one output response, each output response including a sum of: (1) the input drive multiplied by a function of at least one of the at least one modulatory response, each input drive including a function of at least one input, and (2) the recurrent drive multiplied by a function of at least one of the at least one modulatory response, each recurrent drive including a function of the at least one output response, each modulatory response including a function of at least one of (i) the at least one input, (ii) the at least one output response, or (iii) at least one first offset, and computing a readout of the at least one output response.

Claims22

Dependent claims

A method for recursive quadrature filtering an input signal using a recurrent neural network, the method comprising:

receiving, by one or more computational engines, an input drive and a recurrent drive;
computing, by the one or more computational engines, at least one output response of a recursive quadrature filter of at least one input signal, using one or more computational engines, the at least one output response generated by the recurrent neural network depending on:
(1) the input drive comprising a weighted sum of the at least one input signal, and
(2) the recurrent drive comprising a weighted sum of the at least one output response, wherein the weighted sum comprises weights constrained by a preferred temporal frequency;
and
computing a readout of a function of the at least one output response of the recursive quadrature filter of the at least one input signal.

3 more elements

A system for recursive quadrature filtering an input signal using a recurrent neural network, the system comprising:

one or more processors configured to:
receive, by one or more computational engines, an input drive and a recurrent drive;
compute, by the one or more computational engines, at least one output response of a recursive quadrature filter of at least one input signal, using one or more computational engines, the at least one output response generated by the recurrent neural network depending on:
(1) the input drive comprising a weighted sum of the at least one input signal, and
(2) the recurrent drive comprising a weighted sum of the at least one output response, wherein the weighted sum comprises weights constrained by a preferred temporal frequency;
and
compute a readout of a function of the at least one output response of the recursive quadrature filter of the at least one input signal.

4 more elements

Dependencies are inferred from the claim text.

System, method, and apparatus for recurrent neural networks

Life of this patent

System, method, and apparatus for recurrent neural networks

Life of this patent

Abstract

Claims22